THE CHEMISTRY OF GALLIUM

N. N. Greenwood
Department of Chemistry, King's College, Newcastle upon Tyne, England

I. Introduction . . . . . . . . . . . . . . . . . . . Y l
11. Lower Valencies (Oxidation States I and 11) , , , . . . . , , 93
A. Unstable Gaseous Moleculw . . . . . . . . . . . . . Y 3
B. Chalcogenides . . . . . . . . . . . . . . . . . 9 4
C. Halides . . . . . . . . . . . . . . . . . , . 9 6
111. Binary Compounds of Gallium(II1) with Groups V Elements . . . . 99
IV. Binary and Ternary Oxides and Hydroxides . . . . . . . . . 105
V. Compounds with Sulfur, Selenium, and Tellurium . . . . . . . . 103
VI. Gallium Trihalides and Their Addurt,s . . . . . . . . . 105
A. Gallium(II1) Fluoride . . . . . . . . . . . . . . . 105
B. Gallium(II1) Chloride, Bromide, and Iodide . . . . , . , . 106
C. Addition Compounds of the Trihalides . . . . . . . . . . 108
D. Electrical Conductivity and Related Properties . . . . . . . 113
E. Thermochemistry of Gallium Trihalide Adducts . . . . . . . 116
F. Catalytic Properties . . . . . . . . . . . . . . . 120
VII. Other Salts and Complexes of Gallium(II1) . . . . . . . . . . 122
VIII. Gallium Hydrides and Organogallium Compounds . . . . . . 123
References . . . . . . . . . . . . . . . . . 127

I. Introduction

Gallium has several unique properties which have stimulated interest

in its chemistry ever since 1875 when its discovery by Lecoq de Boisbaudran
(I?') confirmed so eloquently his own predictions and those of Mendeleef
for eka-aluminium. The chemistry of gallium was reviewed by Gmelin
(96) in 1936 and in the following year by Einecke (63) but since that time
only one extensive review has appeared (B), though the number of papers
being published annually on this element now exceeds the total number
of papers included in Gmelin's volume. Several general articles have ap-
peared dealing with certain aspects of the history, occurrence, extraction,
purification, analysis, physical and chemical properties, toxicology, or uses
of the element (10, 18, 23, 32, 65, 233, 276, 286).
During the last decade there have been five main areas of interest in
the chemistry of gallium :
(a) the existence of lower valency states in which the oxidation number
of the metal is +1 or f2,
(b) the structure and semiconducting properties of binary compounds
of gallium with phosphorus, arsenic, and antimony, in group V
and sulfur, selenium, and tellurium in group TI,
91
92 N. N. GREENWOOD

(c) the crystal chemistry of the oxides, hydroxides, and related com-
pounds,
(d) the structure and catalytic activity of the trihalides and their
molecular addition compounds,
(e) the preparation and reactions of hydrides and organometallic com-
pounds of gallium.
These topics will be dealt with in successive sections of this review which
also refers to recent work on the salts and coordination complexesof gallium.
The physical constants of gallium which are of interest to chemists are
summarized in Table I. The atomic radius and first ionization potential of
TABLE I
PHYSICAL
PROPERTIES
OF GALLIUM(M.4276)

Atomic number 31 Melting point ("C) 29.75'

Atomic weight 69.72 Boiling point ("C) -2227""
Atomic radius (A) 1.245 d P (solid)(gm/ml) 5.907
Ionic radius (A) 0.62 d P 8 (liquid)(gm/ml) 6.0948
Ionization potential (ev) 6.00 Atomic volume (29")(ml) 11.81
Std. electrode potential, (volts) 0.52 Atomic volume (30")(ml) 11.44

a This figure is obtained by an extrapolation of 1000° from vapor pressures below

1 mm!
gallium are almost identical with those of aluminium and the two elements
frequently resemble each other in chemical properties. Both are amphoteric,
but gallium is less electropositive as indicated by its lower electrode poten-
tial. Differences in the chemistry of the two elements can be related to the
presence of a filled set of 3d orbitals in gallium.
The unusual physical properties of metallic gallium arise from its unique
crystal structure which consists of Gaz units arranged in deformed sets of
hexagonal rings (20, 188). Each gallium atom has one nearest neighbor
a t 2.442 d. The orthorhombic unit ceIl has 4Gaz units lying symmetrically
in the a-c plane at angles of 17"to the c-axis, the spacing of these molecular
planes being b/2. This rather open structure collapses to a more nearly
closepacked arrangement in the liquid state, and melting is accompanied
by a contraction of 3.1% in atomic volume and by a considerable increase
in electrical conductivity (222, 223, 228). Here gallium resembles its neigh-
bor germanium and also antimony and bismuth. The element has a very
low m.p. and an extraordinary low vapor pressure (less than 1 mm at 1300").
These factors combine to give gallium the longest liquid range of any known
substance and form the basis of its (very limited) use in high temperature
thermometers (286). An alloy of gallium, indium, and tin, melting a t 5"
and having excellent wetting properties and a vapor pressure of < lo-' mm
a t 500" has been suggested as a high-vacuum seal (198).
 THE CHEMISTRY O F GALLIUM 93

II. lower Valencies (Oxidation States I and II)

Compounds in which the valency of gallium is less than +3 have been
known since the earliest experiments of de Boisbaudran (96),but it is only
in the last decade that structural studies have elucidated the variety of
ways in which gallium achieves these lower oxidation states. It is now
recognized that gallium forms a range of stable univalent compounds,
many of which also contain an equal number of gallium atoms in the +3
oxidation state, thus leading to a formal over-all oxidation state equivalent
to Ga'"). The +2 state can also be achieved by the presence of Ga-Ga
bonds in the structure and unstable gaseous compounds are known in which
gallium is in a lower valency state.
A. UNSTABLEGASEOUSMOLECULES
Gallium monochloride is formed at 1100' either by the dissociation of
the trichloride or by the reaction of gallium metal with chlorine (191).
GaCla = GaCl + C1,
2Ga + C12 = 2GaCl
Other diatomic halides are formed similarly. Faint absorption bands were
also observed (209) when the arc spectrum of gallium was studied in the
presence of hydrogen and these have been assigned to the compound GaH.
The interatomic distances and energies of dissociation of these molecules
into atoms are recorded in Table 11.
TABLE I1
DIATOMIC
MOLECULES
CONTAINING
Ga")
1
I
Molecule GaI-1 GaF GaCl GaBr GaI

r (R) -1.66 1.775 2.202 2.353 2.575

A&issw. (kcal/mole) -78 143 115 - -
Reference 196 8 6 , 181 6 6

It has been suggested (196) that these diatomic molecules are essentially
ionic, like those of the gaseous alkali metal halides, but the binding energies
calculated from thermochemical cycles do not agree well with those cal-
culated on the basis of pure ionic attraction using the latest experimental
values for the internuclear distances. Moreover, the radius of the Gaf ion
calculated from the observed internuclear distances and the radii of the
gaseous halides ions (231) decreases from 0.74 A, to 0.56 A, the mean value
of 0.63 A being almost identical with the accepted value for Gas+. The
univalent ion would be expected to be considerably larger than this and a
value of 1.17 A has been suggested on the basis of a semiempirical theory
94 N. N. GREENWOOD

(308).On a covalent model, however, the calculated radius for gallium

increases with increasing atomic number of the halogen atom from 1.06 A
to 1.241 8,which is close to the accepted value of 1.245 A for the covalent
radius of gallium. A covalent model is also consistent with the quadrupole
coupling constants of these molecules which have been interpreted in terms
of partial s p hybridization (6).
Flame reactions provide another source of unstable species containing
univalent gallium. Typical reactions are the following (33):
Ga + HzO = GaOH + H
GaOH = Ga + OH

The value for the heat of dissociation of GaOH according to Eq. (4)is
102 rt 5 kcal/mole (33). Similarly, from the partial pressure of gallium in
oxyacetylene or oxyhydrogen flames of known temperature, the dissociation
energy of the molecule GaO was found to be 115.6 f 3 kcal/mole (127).
B. CHALCOGENIDES
Compounds of univalent gallium have been reported for all elements of
group VI except polonium, but little is known of their structure. Gallium(1)
oxide can be prepared by reduction of the sesquioxide with gallium a t 700"
and is purified by vacuum sublimation a t 500" (96):
GarOI + 4Ga = 3Gaz0 (5)
The dark brown, diamagnetic powder is stable in dry air, but undergoes
surface oxidation a t higher temperatures. Above 800" it disproportionates
according to the reverse of Eq. (5). More recently (84) Ga20has been made
by the reaction of gallium with carbon dioxide a t 10 mm pressure and
850" :
2Ga + GOa = Ga10 + CO (6)
Gallium(1) sulfide, like the oxide, cannot be prepared directly from the
elements but is obtained by thermal decomposition of higher sulfides (96).
Thus, when gallium(I1) sulfide is heated a t 1100" for several hours in a
stream of nitrogen, gallium(1) sulfide sublimes quantitatively as green
hexagonal prisms or yellow-green plates (263) [see Eq. (9) below]. The
compound is a phase of variable composition Ga2So.s-1.1and has a hexagonal
lattice similar to that of gallium(I1) sulfide (see below). Above 950" pure
gallium(1) sulfide decomposes to a liquid mixture of gallium(I1) sulfide and
metallic gallium which becomes homogeneous above 1150":
GazS (solid) --f Gas (liquid) + Ga (liquid) * GazS (gas) (7)
In this temperature range the gas above the mixture is monomeric GazS
 THE CHEMISTRY O F GALLIUM 95

and the vapor pressure of the system is given by the relation log p,,, =
7.93 - 9098/T. From this the ('b.p." is 1530°, the heat of vaporization 39.6
kcal/mole and the entropy change 21.8 e.u. (263).
Gallium(1) selenide and the previously unknown gallium(1) telluride
can apparently be made by melting the elements together in evacuated
quartz tubes for prolonged periods; their X-ray powder diagrams have been
recorded and indexed (177).
The divalent compounds Gas, GaSe, and GaTe, can all be prepared
by direct reaction of the elements a t high temperatures. They have recently
been investigated as semiconductor materials, their intrinsic energy gaps
being about three times that of germanium. By contrast, gallium(I1)
oxide has not yet been prepared as a stable solid though bands have been
ascribed to the gaseous, diatomic species GaO (141). Some physical prop-
erties are summarized in Table 111. Crystal structure determinations (128,
TABLE I11
OF Gas, GaSe,
PROPERTIES
PHYSICAL AND GaTe

AE'
(kcal/mole)
I

m.p. color unit cell size (if) elec. optic. Reference

__
Gas 970 f 3" yellow 3.75 a 3.578, c 15.47 - - (228,265')
GaSe 960 10" red brown 5 . 0 3 a 3.735, c 15.887 46.3 47.1 (69, 241, 248,
274)
GaTe 824 f 2" grey - a 23.76, b 4.068 41.5 38.2 (69, ,941, ,949)
c 10.46, 6 4 5 . 4 '

248, 249) have shown that the +2 oxidation state in these compounds
results from a layer lattice arrangement in which the characteristic feature
is the occurrence of Gap4+units, i.e., -SGaGaS, SGaGaS-. This immedi-
ately explains the diamagnetism of the compounds. Within each layer
each gallium is tetrahedrally surrounded by three sulfur atoms and one
gallium atom to which it is covalently bound, and the neighboring multiple
layers are held by van der Waals attraction. Interatomic distances are
Ga-Ga 2.46 A, Ga-S 2.34 A. The selenide and telluride have similar
structures. It is interesting to note that none of the chalcogenides of
gallium(I1) crystallize in the TIS-type structure in which half the thallium
atoms are univalent and half tervalent, i.e., Tl+[Tl"'S2]-. The T1"' are
tetrahedrally surrounded by sulfur and the tetrahedra are linked into
infinite chains by sharing edges (69).
Gallium monosulfide and sesquisulfide both disproportionate a t high
temperatures to give a solid phase of variable composition between Ga4S4.8
and GarS6.z:
96 N. N. GREENWOOD

2Ga& Ga& + $S, above 950"

6GaS E Ga& + Ga2S above 970"

The compound Ga&, which is stable up to 1200" and has a density of

3.82 f 0.01 gm/ml, can also be formed directly from the elements (263).

C. HALIDES
Evidence for univalent halides in the solid state is fragmentary. The
monochloride disproportionates readily into a higher halide and gallium;
e.g., a t 820"gallium metal rapidly volatilizes in a stream of hydrogen chloride
gas to give a brown-black sublimate analyzing as GaC10.g4in which small
drops of gallium metal can be seen (83).The gallium-iodine phase diagram
(44) shows that the limiting composition of the halide phase in contact
with molten gallium is GaIl.rs. When this is repeatedly extracted with dry
benzene a residue of composition GaIl.oeis obtained; the X-ray powder
diagram shows this to be a distinct phase rather than a mixture and the
extrapolated m.p. of the mono-iodide was found to be 271". Other indica-
tions of unstable oxidation states below +2 come from anodic oxidation
of gallium in glacial acetic acid (52) or anhydrous ammonia (51) and from
the solubility of gallium in molten dihalides (43, 46, 47, 78, 275).
Very recently, complexes of gallium(1) chloride and bromide have been
reported (24). They were prepared by allowing volatile ligands (L) such
as dioxane, morpholine, or acetylacetone, to diffuse into a benzene solution
of gallium dihalide in a closed system; disproportionation occurred into
gallium trihalide, which remained in solution, and insoluble complexes of
formula L2GaX. The X-ray crystal structure of [Ga(diox)z]Cl shows it to
contain discrete chloride ions and a disposition of the four oxygen atoms
around the gallous ion which is in the form of a distorted trigonal pyramid
(24).
Gallium forms stable dihalides with chlorine, bromine, and iodine and
it has now been established conclusively that this oxidation number arises
from the existence of equal numbers of gallium atoms in the +1 and +3
oxidation states. Pure gallium dichloride is best prepared (120) by heating
the trichloride with slightly less than the stoichiometric amount of gallium
and then volatilizing the excess of trichloride from the molten mixture. The
last traces of impurity can be removed by fractional freezing to a constant
m.p. of 172.4".A variation of this method (78, I87), which employs a slight
excess of gallium, is less satisfactory. The dichloride has also been prepared
by heating gallium in a stream of dry hydrogen chloride (153)or by reacting
it with the stoichiometric amount of mercurous chloride in a sealed tube
(34). The X-ray crystal structure of the dichloride (82), the Raman spec-
trum (303),electrical conductivity of the fused compound (120), and its
 THE CHEMISTRY O F GALLIUM 97

chemical reactions in benzene solution (3,34)all indicate that it is correctly

formulated as gallous tetrachlorogallate, Ga+[GaC14]-. The compound is
thus an unusual example of the stabilization of an unfamiliar oxidation
state by coordination since the ion which is stabilized (Ga+) is not the one
which is coordinated, [Cl-GaCbI-. There are four Ga+[GaC14]- units
in the orthorhombic cell, space group Pnna, a = 7.29, b = 9.72, c = 9.50 A.
Each gallous ion is surrounded by an irregular dodecahedron of 8 chlorine
atoms from six different GaC14- tetrahedra (82). The Ga-Cl distance
within each tetrahedron is 2.19 A and the crystal radius of the Ga+ ion is
1.33 A (82). This is somewhat larger than expected (cf. p. 93) and is
possibly due to anion-anion contact. The Raman spectrum of the fused
compound (303) showed four fundamentals, as expected for a regular
tetrahedral species, and these corresponded closely with the known spec-
trum of the tetrachlorogallate ion (305):115 cm-' ( Y ~ ,strong, sharp, de-
polarized); 153 cm-' (vq, weak, diffuse, depolarized); 346 cm-1 (vl, very
strong, sharp, polarized) ; 380 cm-1 (vg, very weak, diffuse, depolarized).
Fused gallium dichloride is a typical molten salt and the variation of
its electrical conductivity ( K ohm-'cm-') and dynamic viscosity (7 cp) with
temperature and the corresponding activation energies Ex, E,, are charac-
teristic of a compound in which the cation is smaller than the anion (120).
log K = 1.180 - 784/T; E, = 3.59 kcal/mole
+
log 7 = 1.768 1022/T; Eq = 4.68 kcal/mole
The temperature variation of the density and surface tension have also
been determined (120): dqt = 2.4206 - 1.36 X 10-a(t - 170) gm/ml;
7 = 56.6 - 0.18(2 - 170) dyne/cm. Above about 200" gallium dichloride
disproportionates reversibly to gallium and the trichloride and above 400"
the vapor density indicates considerable quantities of monomeric GaClz
molecules (187).
Gallium dichloride is readily soluble in benzene (187),a saturated solu-
tion at room temperature containing 4.60 gm/100 gm solvent (153). The
Raman spectrum of the solution confirms the ionic nature of the solute
(196)and a crystalline phase [Ga-C6H6]+[GaCl4]- can be separated from the
solution (238). Gallous tetrachloroaluminate GaAlCl4 (46),forms a similar
complex (238). Further proof of the two oxidation states of gallium in the
dichloride comes from the reaction of a benzene solution with hydrogen
sulfide: Ga' was precipitated as the sulfide and gallium trichloride remained
in solution (34).
The strongly reducing nature of the dichloride, which was observed by
the earliest workers (231) has been repeatedly confirmed and it has been
shown that carbon tetrachloride is reduced to carbon (187). Limited stabil-
ization of the gallous ion has been achieved by coordination, and a wide
98 N. N. GREENWOOD

variety of complexes of the form [GaL4]+[GaCl&have been prepared with

oxygen, sulfur, selenium, nitrogen, and arsenic donors where L4 represents
four monodentate ligands, two bidentate ligands, or one quadridentate
ligand (3, 24). Such complexes are usually white or pale cream though the
adducts with %,a'-dipyridyl and dithizone were red. Molecular-weight
determinations, electrical conductivity, and diamagnetism confirm the
general ionic formulation but the problem of the detailed stereochemistry
of the [GaL4]+cations has not yet been settled. The gallous ion in such
complexes has five pairs of electrons and is therefore isoelectronic with
phosphorus, sulfur, and chlorine in their compounds PFS, SF4, and C1F3,
all of which are based on trigonal bipyramidal symmetry. If, however, the
4s2 electrons are stereochemically inert, then presumably either planar or
tetrahedral symmetries are possible though these would involve the use
of unusual hybrids. The stereochemistry of the complex [Ga(diox)z]Clhas
already been mentioned on p. 96.
Gallium dibromide is very similar in its structure and properties to the
dichloride. It is best prepared pure by heating the tribromide in a sealed,
evacuated tube at 180" with slightly less than the theoretical weight of
gallium, after which excess of the tribromide is volatilized away and the
compound fractionally crystallized from the melt to a constant melting
point of 166.7' (124, 304). Reaction of gallium with mercuric bromide also
affords gallium dibromide and when this reaction is carried out in the pres-
ence of benzene, the solvate GasBr4.CaHais obtained ($4).
Crystalline gallium dibromide (304) like the dichloride (171) is diamag-
netic, and the Raman spectrum of the fused compound showed only the
four fundamental modes expected for the tetrahedral GaBr4- ion (304) :
79 cm-l (vZ, strong, sharp, depolarized); 107 cm-1 (~4,strong, sharp, de-
polarized) ; 209 cm-' ( Y ~ very
, strong, sharp, strongly polarized); 288 cm-I
(va, weak, diffuse, depolarized). The electrical conductivity ( K ohm-'crn-')
viscosity, (7 cp) and density are given by the equations:
log K = 1.142 - (865/T)
+
log 7 = 1.793 (1158/T)
d = 3.4656 - 1.69 X (1 - 170).
These data, together with the activation energies (E, = 3.96, E,, = 5.30
kcal/mole) show that the compound is a completely dimociated molten
salt in which the cation is small compared with the anion (120, 304).
Gallium dibromide apparently has two crystalline modifications, LY
and 8, melting at 153" and 165" (45). As mentioned on p. 96 the fused
compound dissolves metallic gallium to some extent but complete reduc-
tion to the monobromide could not be effected, the limiting composition
being 31.4 mole % GaBr at 170". In the presence of aluminium tribromide,
 THE CHEMISTRY OF GALLIUM 99

however, complete reduction to monovalent gallium occurred and the com-

pound Ga+[A1Br4]-(m.p. 159") was found to be dimorphic, each modifica-
tion being isomorphous with the corresponding form of Ga+[GaBr4]- (43).
A variety of complexes [GaL4]+[GaBr,]- analogous to the chlorogallates
has also been described (3, 24).
Gallium di-iodide (m.p. 211') can be prepared by direct reaction of the
correct amounts of gallium and iodine (44) or by the reaction of gallium
with mercuric iodide in a sealed tube (34). It is diamagnetic (44),and in the
absence of further data can be presumed to have the structure Ga+GaIa-.
When heated at 250" the di-iodide loses gallium tri-iodide by dispropor-
tionation, leaving a residue of limiting composition close to the monoiodide
(see p. 96).
Ill. Binary Compounds of Gallium( Ill) with Group V Elements
The compounds which gallium forms with nitrogen, phosphorus, arsenic,
and antimony are isoelectronic with the Group I V elements and there has
been considerable interest, particularly in the physical properties of these
compounds, since 1952 when Welker first showed that they had semi-
conducting properties analogous to those of silicon and germanium (287).
Gallium nitride is the only member of the group which cannot be pre-
pared by direct reaction of the elements. It was first made (155) by the
reaction of gallium with ammonia a t 900"-1000" and has also been prepared
by the decomposition of (N€11)3GaFba t 600" (131). Crystalline gallium
nitride is formed when the adduct GaC13.NH3is pyrolyzed a t 1000" (219).
The compound is grey or yellow (131, 155, 192, 229), diamagnetic (131),
and crystallizes in the Wurtzite (ZnS) structure (156).Its physical proper-
ties are summarized in Table I V and the heat of formation -AHto is
24.9 f 0.9 kcal/mole (131).
Gallium nitride is not hydrolyzed by hot water or by acids such as dilute
or concentrated hydrofluoric, hydrochloric, or nitric (155, 229). Aquaregia
is also without effect but the compound dissolves slowly in hot concentrated
sulfuric acid (155). With hot concentrated aqueous alkali, ammonia is
evolved (229).There is no reaction with hydrogen a t 800"but oxygen slowly
reacts with gallium nitride a t 900" ovcr a period of days to give nitrogen
and the sesquioxide Ga203( 1 5 6 ) . The reaction proceeds with ignition in the
presence of paraffin (131).
Although gallium nitride cannot be formed directly from gallium and
nitrogen, the compound Li3Gs reacts wit,h nitrogen a t 600" to give the ter-
nary nitride, Li3GaNZ(158). This light-grey powder, d 3.35, can also be
prepared by heating lithium nitride and gallium in nitrogen a t 730" or by
heating a mixture of lithium nitride and gallium nitride. Lithium gallium
nitride is very similar to lithium aluminium nitride, Li3AlN2:both crystal-
 TABLE IV
PHYSICAL
PROPERTIES
OF 111-V COMPOUNDS

Property GaN Gap GaAs GaSb

m.p. d > 1250" (156,229) 1350" (258) 1238"" (19, 176) 7120b (9, 95)
Crystal structure Wiirtzite (156,157) zinc blende zinc blende zinc blende
a = 3.180, c = 5.166 A a0 = 5.4505 d (90) a0 = 5.6534A (90,210) a0 = 6.0954A (90,175) Z
c/a = 1.625" 5.436 d (74) 5.6560 d (175) 6.1342 d (15) Z
Refractive index - 3.37 (75), 2.9 (258) 3.4 (212),3.2 (258) 3.9 @ I S ) , 3.7 (258)
0
K ohm-lcm-1 10-9-101 (229) 10-e10+* (75) 10*1O+J (134) 12 (567,6-13 (257)
2.5 X 10- (159) (0.17-1.0) X loJ (175) 10-12 (190) Em
AE ev (optical) 3.25 (163) 2.24d (74) 1.25 (50),1.35" (212) 0.68 (Ha),0.71' (50) 2
AE ev (electrical) - - 1.320 (175) 0.7-0.8g $
0
0
t7
Also 1280" (50).
702" (288), 703" (176),705.9' (101), 725" (161).
cA l soa = 3 . 1 6 0 d , c = 5 . 1 2 5 ~ , c / a = 1 . 6 2 2 ( 1 9 2 ) .
d A E = 2.4 - 5.4 X 10-'T (211).
= 1.53 - 5.0 X 10-T (811).
'A E = (0.695-0.704) - 2.9 X lO-lT (832).
n0.71 (175, 190),0.775 (15),0.78 (55),0.80 (IN)0.82 , (212,288).
 T H E CHEMISTRY O F GALLIUM 101

lize in the antifluorite structure and are readily hydrolyzed by water or

dilute acids (158). Ternary phosphides and arsenides can also be prepared
by heating mixtures of the binary compounds a t 450-650'. Lithium gallium
phosphide, Li3GaP2,resembles the nitride in having the antifluorite struc-
ture but the arsenide, LirGaAsz, has a tetragonal unit cell with the axial
ratio c/ a = 1.016 (160).
Gallium phosphide, arsenide, and antimonide can all be prepared by
direct reaction of the elements; this is normally done in sealed silica tubes
(19, 210) or in a graphite crucible under hydrogen (9, 190, 257). Phase
diagrams indicate the stoichiometry GaAs and GaSb (176). Similar data are
hard to obtain in the gallium-phosphorus system because gallium phosphide
has a vapor pressure of more than 13.5 atm a t its melting point (73)-cf.
0.89 atm for gallium arsenide (73).
The physical properties of these three compounds are compared with
those of the nitride in Table IV. All three adopt the zinc blende crystal
structure and are more highly conducting than gallium nitride. Whereas
the group IV semiconductors silicon and germanium are purely covalent,
the 111-V compounds have some ionic character and provide a wider range
of basic semiconductor parameters such as energy gaps and mobilities.
Estimates of the precise extent of heteropolarity vary widely as is to be
expected, but, independently of the method used to determine ionic charac-
ter (221, S o l ) , there is general agreement that this diminishes in the
sequence GaP > GaAs > GaSb. The energy gap L\E decreases in the same
order (see Table IV). In general, the m.p. and energy gap of 111-V com-
pounds are higher than those of the isoelectronic Group IV elements,
despite the similarity in bond distances.
Gallium phosphide forms orange, transparent crystals and is potentially
useful for very high temperature work particularly where high frequency
performance is not required. Gallium arsenide forms good pn-junctions
and has been used to study photovoltaic effects. It is one of the most
promising semiconductors for high frequency and high temperature work.
Gallium antimonide is the most studied compound of the group because
of its low melting point. It is normally a p-type semiconductor but n-type
samples have been prepared by doping with tellurium.
Further flexibility is introduced by the possibility of alloying these
various compounds. There is normally a complete range of solid solutions
with other 111-V compounds (306, $Or), and properties tend to vary

w
approximate1 linearly with composition (14,176). By contrast, germaniup
(a0 = 5.658 ) is less than 2% soluble in gallium arsenide (a0 = 5.656 A)
despite the near identity of their lattice parameters (161).
The chemical properties of this group of compounds have been less
studied. There is no reaction with water or di1ute:alkali but dilute hydro-
102 N. N. GREENWOOD

chloric acid yields small quantities of arsine or stibine (161). Ternary

compounds such as LiaGaPzand Li3GaAszcan be formed by direct reaction
of lithium phosphide or arsenide with gallium phosphide or arsenide (160).
IV. Binary and Ternary Oxides and Hydroxides
The precipitation of the hydrous oxide of gallium from aqueous solutions
and its dissolution in excess of alkali has been a subject of continuing
interest (144, 150, 270, 284). For example precipitation from an 0.007M
solution of gallium trichloride begins a t pH 3.15 and is complete a t pH 6 ,
after which redissolution begins and is complete a t pH 9.5 (183). The
solubility product [Ga3+].[OH-]3is 10-36.6 and may be compared with those
of aluminium and indium hydroxides which are and 10-33*8(183).
In the presence of sulfate ions the hydrous oxide begins to precipitate after
approximately one OH- per mole of gallium has been added, but aqueous
solutions of the chloride, bromide, or nitrate can be treated with up to three
OH- per mole of gallium without precipitation though addition of more
hydroxyl ions causes the solutions to flocculate sharply. Diffusion current
data show that this unusual behavior is due to the formation of Ga3+-OH-
aggregates rather than to peptization (203) and detailed isopiestic studies
on perchlorate solutions suggest that the degree of polarization of the oxy-
gen-bridged cationic gallium aggregates increases from 1.3 a t a mole ratio
OH-/Ga3+ of 0.5, to 84 a t a mole ratio of 2.0 (216). Further work using
Hayleigh turbidities and refractive index increments indicates that Sillen's
core-link hypothesis applies to these aggregates (239).
When freshly precipitated gallium hydroxide is dissolved in saturated
aqueous lithium hydroxide and the solution evaporated, the compound
[Li(Hz0)4][Ga(OH)4] crystallizes. This readily loses water over sulfuric
acid a t room temperature, and again a t 100" to give compounds which
can be formulated as [Li(H20)2][Ga(OH)4] and Li[GaO(OH)z]. Further
heating results in irreversible dehydration to LiGaOz (146). Sodium
hydroxide yields Na[Ca(OH)r] and NaGaOz and potassium hydroxide
yields K[Ca(OH)4],KGaOz-l$HzO,and KGaOZ-H20.By contrast, calcium
and strontium hydroxides tend to give six-coordinate complexes of the
form Ma Ga(OH)6]2(145). The phase diagram of the system GanOa-NazO-
HzO has also been studied and various hydrated gallates established (89).
(See also ref. 79).
Gallia, like alumina, exists in a variety of structural forms and these
have been extensively investigated by electron diffraction and X-ray
crystallographic methods. Forms designated a, p, y, 6, and r have been
distinguished, the stable form between room temperature and the m.p.
being &Ga203which is isomorphous with 8-Al2O3 (79, 174, 235). The low-
temperature thermodynamic properties of p-Gaz03 have been tabulated
(1, 167) and its m.p. is 1725' (139).The hydrous oxide, GaO-OH, has the
 THE CHEMISTRY OF GALLIUM 103

same structure as diaspore, p-AlO-OH (16), and undergoes dehydration to

@-Gaz03 a t 300" (185, 235). A metastable phase Ga(OH), and possibly also
GaZO(OH), have also been recognized (185). Treatment of the first of
these compounds with calcium hydride removes only one mole of water and
this has been taken to indicate that its structure is (;aO.OH*HzO rather
than Ga(OH)3 ( 5 ) .
The close similarity in ionic radii between gallium and aluminium which
leads to the similarity in oxide and hydrous oxide phases just mentioned,
also ensures that the two series of compounds exhibit extensive solid solu-
tion. Thus, below 810°, the compound C;aAIOJand a series of solid solutions
extending almost to pure alumina and gallia are stable. When water is
added, these phases are stable only above a transition temperature in the
region of 300400" (depending on composition). Below this temperature
a complete series of diaspore-type structures extends from A10.OH to
GaOSOH and below about 275' a series of boehmite-type structures extends
from AlOSOH to a composition in which 30% of the aluminium ions have
been replaced by gallium (139, 236). Similarly gallium analogs of the
aluminosilicates and aluminogermanates have been prepared (7 ). For
example, fully substituted or partly substituted counterparts have been
prepared of sodium feldspar, (NaA1Si308), calcium feldspar, orthoclase
(k'AlSi,Os), leucite (KAlSi206), anorthite (CaAlzSi208), and mullite
(3Al2O3.2SiO2)(86, 97, 189). Both silicon and aluminium can be com-
pletely removed without change in structure from gehlenite (Ca2AlzSi07)
+ +
by replacing aluminium by gallium and (Si Ca) by (Ga Ln) to give
CaLnGa207 where Ln is a lanthanide element (59).
Similar isomorphous substitutions in the sesquioxide (Gaz03)itself lead
to perovskite lattices such as LaC:aOB,CeGaOs, NdGa0, (165, 234) and, as
expected, substitution can also occur in spinel-type lattices, e.g., half the
aluminium ions in chrysoberyl (+anbe replaced to give BeAIGaOI (93).
A variation on the III-V theme discussed in the preceding section con-
cerns the replacement of pairs of silicon atoms in silica by gallium and a
group V element. Thus CaP0, exists in forms corresponding to a low-
temperature quartz, a metastable, low-temperature crystobalite and a
high-crystobalite with transition temperatures a t 933" and 616" respectively
(205, 217, 2'51). The compound GaAsOI exists only in the low-temperature
quartz form and with the compound GaSb04, where the increased cation
radius requires 6 :3 rather than 4 :2 coordination, a disordered rutile
(TiOz-type) structure is adopted (21, 251).
V. Compounds with Sulfur, Selenium, a n d Tellurium
The sesqui-sulfide, -selenide, and -telluride of gallium can all be pre-
pared by direct reaction of the elements a t high temperature. The com-
pounds are isomorphous with zinc sulfide (132) and this, rather than iso-
104 N. N. GREENWOOD

morphism of zinc sulfide with gallium arsenide, is now thought to be the

reason for specific enrichment of gallium in sphalerite minerals (99).
Gallium sulfide has three crystal modifications (128, 132) :

yGazS3(zinc blende) a. = 5.171 f 0.004 8,stable at low temperatures; the

gallium atoms randomly occupy 2/3 of the tetrahedral sites in the sulfide
cubic close-packed lattice.
j3GazS3(Wurtzite) a = 3.678 f 0.005, c = 6.016, c/a = 1.636;the gallium
atoms randomly occupy 2/3 of the tetrahedral sites in the sulphide
hexagonal close-packed lattice; y j3 transition 550"-600".
aGazS3(Wurtzite superlattice with unit cell containing 6GazS3);a = 6.370;
c = 18.05 A, C / U = 2.833.

Gallium selenide and telluride exist only in the zinc blende structure with
two-thirds of the cation sites occupied by gallium atoms. Unit cell dimen-
sions are: GazSe3a. = 5.430 f 0.005 8 (99) [or 5.418 8 (132)l; GazTe3
a. = 5.887 f 0.003 8 (99) [or 5.874 8 (132)l. In all cases bond distances
are close to those expected.
All three compounds are semiconductors, though their use is somewhat
restricted by their ready hydrolysis. Activation energies are intermediate
between those of the isoelectronic II-VI and III-V compounds (98), typical
values for the energy gap AE between bonds being: GazS3 2.85 ev (164);
GazSe31.9 ev (164); GazTe31.0-1.55 ev (136, 137, 138, 164). The energy
gap decreases with decreasing ionic character (cf. Gaz03 4.4 ev) and the
above values are thus less than those of the corresponding aluminium chal-
cogenides but greater than those of the indium analogs (164). Ternary
compounds such as CuGaSz, AgGaSz, and the corresponding selenium and
tellurium compounds which all adopt the chalcopyrite structure (130)
have also been investigated as semiconductors (4, 71, 94). The compounds
CdGazSa,HgGazS4,CdGazSe4,IIgGaBer, and CdGazTer have been prepared
by direct reaction of the component binary chalcogenides and all adopt a
chalcopyrite structure in which the appropriate number of cation sites is
left vacant (129). The corresponding reaction between cuprous oxide and
gallium oxide leads to the compound CuGaOz, but this, being more ionic
than the chalcogenides, crystallizes in the NaHFz-type structure with octa-
hedral, rather than tetrahedral coordination (133).
An interesting series of alkali metal thiogallates MzGazSahas been
prepared (146) in which the anion is considered to have the structure:
 THE CHEMISTRY OF GALLIUM 105

The lithium and sodium compounds are prepared by heating an equimolar

mixture of alkali metal carbonates and gallium sesquioxide to 900' in a
stream of hydrogen sulfide whereas the potassium, rubidium, and cesium
salts are prepared by heating the carbonate and oxide in molten sulfur
and then leaching with water to leave the insoluble thiogallate. The five
compounds are yellow or light brown, have m.p.'s in the range 952-1020',
are stable in air or water, but are decomposed by strong acids with libera-
tion of hydrogen sulfide (146).
VI. Gallium Trihalides and Their Adducts
Current interest in the trihalides of gallium centers on their structure
and catalytic activity, and on the stoichiometry, structure and stability
of their coordination complexes. Gallium fluoride differs markedly from
the other halides and in this it resembles the fluorides of aluminium and
indium. For example, the m.p.'s of the fluorides are several hundred degrees
above those of the chlorides, though this difference decreases from 1097"
for aluminium and 872" for gallium to 584' for indium.
A. GALLIUM(III) FLUORIDE
Gallium trifluoride can be prepared by dissolving gallium hydroxide in
40% hydrofluoric acid and adding this to a saturated solution of ammonium
fluoride. The complex (NH4)3GaF6so obtained is heated to 630' in a current
of fluorine to yield the trifluoride (135). More recently it has been made by
direct reaction of hydrogen fluoride and gallium at 440' (26). This method
is to be preferred since samples prepared from the hexafluorogallate are
invariably contaminated with nitride derivatives. When the hexafluoro-
gallate is heated at 220' in a stream of nitrogen the compound NHdGaF4 is
obtained. This has tetragonal crystal symmetry with a = 3.71 A, b = 6.39 A
and is isostructural with several tetrafluoroaluminates (26). Gallium
trifluoride is isostructural with ferric fluoride and the rhombohedra1
unit cell has a = 5.20 A, 0 = 57.5' (25). The density is 4.47 and the com-
pound is said to have a m.p. greater than 1000" and to sublime readily a t
950" (135). An electron diffraction study of the gas at high temperature
indicates the presence of GaFa molecules in which the Ga-F distance is
1.88 A and the F-F distance 3.25 A; dimeric molecules, GazFe,if present,
have a concentration of less than 1% (2).
The hexafluorogallate mentioned in the preceding paragraph presum-
ably contains 6-coordinate gallium and there is strong presumptive evi-
dence for sixfold coordination in the following complexes : Li3GaF6,
Na3GaF6,K2GaFs.Hz0,RbGaF4.2H20, CsGaF4.2Hz0, (224), as well as in
GaF3.3Hz0 and GaF3.3NH3 (169). A series of complexes [M(HzO)e]-
[GaFs.HzO](M = Mn, Co, Ni, Cu, Zn, Cd) has also been described and
106 N. N. GREENWOOD

also Ba3[GaFs]z.Hz0,3SrFz.GaF3.3H20,Ag,GaF6.10Hz0, and T12[GaF6.H~0]

(225).
The trihydrate of gallium trifluoride is formed by dissolving gallium or
its nitrate or oxide in 40% hydrofluoric acid and evaporating the solution
(169, 271). The triammine cannot be formed from the trifluoride directly
but is prepared via the trihydrate (169). The affinity of the gallium ion for
fluoride ions has been measured spectrophotometrically and has been
found to be greater than that of chromium for fluoride. At 25" and an ionic
strength of 0.5 the quotient [GaFz+][H+]/([Ga3+][HF]) is 120 compared
with 26 for the corresponding Cr3+system (300).
B. GALLIUM(III)CHLORIDE, BROMIDE, AND IODIDE

Gallium trichloride is readily prepared in quantitative yield by the

reaction of either chlorine or hydrogen chloride gas on heated gallium
metal (96, 113, 154). Chlorination may also be effected by heavy metal
chlorides such as AgCl, HgC12, or PbCl2; in these reactions, which have
potential use for preparing chlorine-36 labeled gallium trichloride, an
equilibrium is established and reasonable yields can only be obtained by
volatilizing the chloride from the reaction zone (37). Gallium trichloride
is normally purified by vacuum sublimation, though zone refining has also
been suggested (218, 230).
The Raman spectrum of gallium trichloride has been interpreted in
terms of bridged dimeric molecules GazC16(87) and this is also consistent
with the very low electrical conductivity of the melt (119). When the com-
pound solidifies, however, the conductivity increases ninefold to a value
comparable with that of a fully ionized salt near its m.p. It is concluded
that the m.p. of gallium trichloride is the transition temperature between
an ionic solid and a covalent liquid (119). A similar phenomenon has been
observed for gallium tribromide (119) aluminium trichloride ( I S ) , and
phosphorus pentachloride (256).
Some physical properties of gallium trichloride are given in Table V.
The low m.p. is particularly noteworthy. The vapor density of the gas
shows it to be dimeric with a degree of dissociation of 0.4% a t the m.p.
and 2% at the b.p. (72). The complexes of gallium trichloride are discussed
in Sections VI C, D, and E.
Gallium tribromide is readily prepared by the reaction of heated gallium
with a bromine-nitrogen gas mixture (72, 170) and detailed experimental
procedures have been described (123). The compound can also be prepared
from gallium and hydrogen bromide (26).
Like the trichloride, solid gallium tribromide has an appreciable elec-
trical conductivity which decreases by a factor of 23 when the compound
is fused (119). Other properties of the tribromide are summarized in Table
 TABLE V
PHYSICAL ~ S GALLIUM
P R O P E R TOF TRIHALIDES

Property GaCla
~ --
GaBrr Ga13
~.
s
M
m.p. Ti.i5 f 0.05" (lf5) 122.3 5 0.05" (119) 211.5 rt 0 . lo,211" (44,119)
b.p. 201.3" (7.9) 279O (72) 346" (72)
Pbm) 10.4 (at 78") (72, 187) 4 . 7 (at 125") (72) 19.2 (at, 215") (7%)
d4s (solid) 2.47 (170) 3.69 (170) 4.15 (170)
drr (liquid) 2.05314).00200(t- 78) (115) 3.10764.002471't - 125j (121) 3.5914.00224(t - 215) (170)
v (CP) 1.816 (at i 8 " ) (115) 2.780 (at, 125") (121) -
E,(kcal/mole) 3.64 (115) 3.94 (121) -
r(dyne/cm) 27.3 - 0.105(t - 78) (115) 34.8 - 0.15(1 - 125) (121) -
K (ohrn-lcm-') 1.86 X (at 78") (119) 7 . 2 X 10' (at 125") (119) 1.18 X lov4 (at 212') (119)
AH,.,(kcal/mole) 11.8, 12.3 (72, 187) 14.3 (72) 18.1 (7%)
AH~,,(kcal/mole) 5.2 (187) - AHeubi23.0 (259)

*
Affdi,.,(kcal/moIe) 21.0 (per Ga2Cls) (72) 18.5 (per GazBre) (75') 11 . 0 (per Ga&) (78)
AH"r(kcal/mole) 125 f 1 (168) 92.4 0 . 3 (168) (-61) (168)
108 N. N. GREENWOOD

V from which it is seen that there are trends to increasing m.p., density,
and viscosity and to decreasing volatility, heat of dimerization, and heat
of formation with increasing atomic weight of the halogen.
Gallium tri-iodide is prepared by direct reaction of the elements in a
sealed tube (96, 119). It is important to avoid excess of iodine since this
cannot easily be removed by vacuum sublimation of the compound (44).
The trends noted in Table V for the chloride and bromide continue for the
iodide. I n addition the degree of dissociation of the saturated vapor into
monomeric GaX3 units a t the m.p. is 0.4% for the chloride, 14% for the
bromide and 96% for the iodide. The corresponding figures a t the respective
b.p. are 2%, 30%, and 8701, (7'3). The virtually complete dissociation of
the tri-iodide into monomeric molecules is confirmed by electron diffrac-
tion on the gas which indicates the predominance of planar Ga13units (26)
as in GaF3 (2). Structural deductions have also been made from nuclear
quadlupole measurements on crystals of the trihalides of gallium (35).
C. ADDITIONCOMPOUNDS OF THE TRIHALIDES

A discussion of the molecular addition compounds of the gallium

trihalides must include reference to their stoichiometry, structure, and
stability. All these aspects have been studied extensively during the last
decade, and a fairly complete picture is now beginning to emerge. The
known complexes of gallium trichloride are listed in Table VI from which
it is seen that 1:1 complexes predominate though some compounds are
well established in which there are two moles of ligand, and alkyl chlorides,
in addition to forming 1:1 complexes, apparently form complexes in which
one more of ligand is associated with two GaC4 units. Intercalation com-
pounds of gallium trichloride in graphite have also been described (237')
but these nonstoichiometric adducts are of a different type and will not be
considered further. Fewer complexes of gallium tribromide have been
described (see Table VII) and the coordination chemistry of the tri-iodide
is almost completely unexplored.
The general properties of complexes with organic ligands will be men-
tioned first, and this will be followed by a discussion of the complexes
with inorganic ligands. Aspects which relate specifically to the electrochem-
istry or thermochemistry of these systems will be deferred until Sections
D and E.
Gallium trichloride resembles aluminium chloride in forming no com-
pound with anhydrous hydrogen chloride. However, in the presence of an
aromatic hydrocarbon such as toluene a compound is formed having the
structure [PhMeH+][GaC14-] (68).Ethers can also be used to solvate the
proton (299).
Organic nitrogen-containing ligands, as expected, form strong com-
 TABLE V I
COMPLEXES
OF GALLIUM TRICHLORIDE

Complex m.p. Reference I Complex m.p. Reference

GaCkMePhHCl - (28) GaCL MeCl 48"

GaClr MeaN >200° (283) 2GaCkMeCl -
GaC4.2MeaN d > -48" (283) GaCkEtCI -
GaClaC6H6N 126" (108,116,148) 2GaCL.EtCl (-29" calc)
GaC&.2C&N 113" (108,116,148) GaClZ.Pr'C1 -
GaCl&Hl0NH 134' (108,117) 2GaCb-Pr'C1 -
GaCL.2C6HloNH 112" (108,117) GaCk PhaCC1 -
Gaa-3phen0 - 11.47) CaC1rNHab 124"
GaCh*PhCN 125" (280) GaCk NOCl -
GaC&.PrCN - (280) GaCb.PCl3 28" incongr.
GaC4-PhNO2 64" (68) GaCb PCL 368-371"
GaC&-p-MeCeH4NOz 95" (250) GaCk POCh 118.5'
GaCkMerO 3.0" (282,283) GaC13,AsCls incongr.
GaCb-EkO 16.2" (102,105) GaC13AsCbCl2 5"
GaCL-2EbO 9" (102,105) GaCkLiCl -
GaCkEtpOHCl - (299) GaCkKCl 259"
GaCL.Me&O 42 .ao(d) (106) GaCL-CsCl 385"
GaC4-MeCOCl 86"( d ) (42, 106,11.3) GaCb.NH4Cl 304"
GaC4-PhCOCl 46.5" (113,280) GaClz-GaC1 172.4'

phen = o-phenanthroline, i.e.,

Also at ratios of 3, 5, 6, 7, and 14 moles of NHa per mole of GaCh (172). c

0
co
110 N. N. GREENWOOD

TABLE VII
COMPLEXES
OF GALLIUMTRIRROMIDE
Complex m.p. References Complex m.p. References

>200° GaBrs.NHpb
d>-48' GaBra.POCla
126" GaBrs.POBr8
90-100° GaBmKBr
139' GaBrsCsBr
- GaBmEtlNBr
8.5" GaBra-GaBr

a phen = o-phenanthroline, i.e.,

Also at ratios of 5, 7, 9, and 14 moles of NH: per mole of GaBrs (172).

plexes. The 1:1 complexes of trimethylamine with gallium trichloride and

tribromide are chalky powders stable in anhydrous conditions to 200"
(283). Complexes with two moles of trimethylamine also form a t low tem-
peratures but these dissociate a t temperatures above -48" to give the 1: 1
compound and trimethylamine (283). Conductimetric behavior of these
systems in nonaqueous solvents have been investigated. Thus addition of
a ligand such as dimethylamine, dimethyl ether, or ammonia to a solution
of gallium trichloride or tribromide in nitrobenzene rapidly increased the
conductivity of the system, and this was followed by a slow increase over
a period of up to two days before an equilibrium value was attained (282).
Detailed interpretation is further complicated by the fact that gallium
trichloride and nitrobenzene themselves form a complex, m.p. 64", which
can dissociate ionically, the specific conductivity of the fused compound
being 5.8 X ohm-km-l a t 75" (68).
Pyridine and piperidine also form 1:1 and 2: 1 adducts with gallium
trichloride, (108, 116, 117, I @ ) , but only 1:l complexes with the tri-
bromide. The complexes with two moles of ligand are both relatively
unstable and evolve 1mole of ligand when warmed under vacuum. Evidence
for the compounds GaBr3.3C6H6Nand G & I ~ * ~ C S & inNwhich the gallium
is presumably six-coordinate has also been adduced (48). The tris(orlho-
phenanthroline) complex appears to be of a different type and has been
formulated as a six-coordinate trisbidentate complex of the Ga3+cation,
[Ga~hen3~fI [C1-I3 (147).
Little is known about the complexes of gallium trichloride with benzo-
nitrile and butyronitrile though the dipole moment of the former in benzene
solution has been found to be 8.65 D (2880).The corresponding value for
 THE CHEMISTRY OF GALLIUM 111

p-nitrotoluene-gallium trichloride is 9.16 D (280). A cryoscopic study of

the nitrobenzene-gallium trichloride system in benzene revealed an un-
stable incongruently melting compound GaC13.2PhN02 but the stable
adduct is again the 1: 1 complex (68).
Ethers readily complex with the gallium trihalides, and there seems to
be less tendency than with the aluminium analogues for thermal elimina-
tion of alkyl halides. Gallium trichloride and tribromide form 1 : l com-
plexes with dimethyl ether which melt below room temperature (285).
Vapor pressure-composition rurves at -82" show only the 1: 1 complex
but in nitrobenzene solution there is some indication of the attachment
of a second mole of ligand (283). This behavior contrasts with that of
diethyl ether since the phase diagram of this ligand with gallium trichloride
shows two well defined complexes melting a t 16.2' and 9" (105). The
bis(diethy1 ether) complex reverts to the 1: 1 complex in vacuo a t room
temperature. Vapor pressure data on some of these complexes are as fol-
lows (110, 283):
GaC13-MenO: log p,, = 6.52 - 1956/T; AH,,, = 8.95 kcal/mole;
p(43") = 2 . l m m
GaBra.Me20: log p,, = 4.61 - 1205/T; AH,,, = 5.51 kcal/mole;
~ ( 2 6 " )= 3 . 4 m m
GaC13.EtzO: log p,,, = 9.56 - 2780/T; AHvap= 12.72 kcal/mole;
p(26") = 1 . 8 mm
The relatively high vapor pressure and low heat of vaporization of the
bromide complex is noteworthy.
The only reported ketone complex of a gallium trihalide is gallium tri-
chloride-acetone which has been established by phase studies; it melts with
some decomposition a t 42.2" (106). This contrasts with the many known
complexes of aluminium halides with ketones and reflects the limited study
which has been afforded to the gallium systems rather than their instability.
Acetyl and benzoyl chlorides form 1:1 adducts with gallium trichloride
and the chemical (113) eIectrica1 (113) and infrared spectroscopic (42) evi-
dence suggests that the structure is predominantly of the form [RCO+]-
[GaCL-]. In benzene solution the dipole moment of the compound GaCla.
PhCOCl is 6.85 D (280). Phase studies indicate the absence of adducts a t
stoichiometric ratios other than the 1 : 1 (113).
Alkyl halides form weak complexes with gallium trichloride a t low
temperatures (27, 302). In addition to the 1 : l complexes, solvates of the
type IiC1.GazCls were also observed though their structure has not been
established. The heat of dissociation of the complex EtC1.2GaCls was fouiid
to be 0.74 kcal/mole (302) arid dissociation pressure measurements on the
corresponding system with methyl chloride led to the following heats (261):
112 N. N. GREENWOOD

2(MeCbGaCla) (c) = MeC1.2GaCla (c) + MeCl (g); AH = 7 . 1 kcal

MeC12GaCla (c) = 2GaClg (c) + MeCl (g); AH = 8.1 kcal
MeCI.GaC13 (c) = GaCla (c) + MeCl (g); AH = 7 . 6 kcal

The 1:1 complex is therefore moderately stable. The slow rate of halogen
exchange between gallium trichloride and methyl bromide at -80"suggests
that the 1:1 complexes are not ionized as R+GaC14- but should be formu-
lated as covalent adducts RX+GaX3 (27).The solutions are unstable and
evolve hydrogen halide below room temperature, the rate increasing with
increasing chain length of the alkyl group (302). Isomerization of Pr"C1
to PrCl has also been noted. Other compounds such as GaBr3-EtBrhave
been postulated to explain kinetic results but have not been isolated (261).
The compound Ph3C+GaC14-has been prepared as a yellow solid by mixing
benzene solutions of gallium trichloride and triphenylmethyl chloride (102).
Ammonia forms a variety of compounds with the gallium halides of
empirical formula GaX3.nNH3though their structure is unknown (172). It
is not clear that all are simple adducts or solvates, and solvolysis may well
have occurred at the higher mole ratios. Values of n are 1, 3, 5, 6, 7, and
14 for gallium trichloride, 1,5,7,9, and 14 for the tribromide, and 1,5,6,7,
9, 13, and 20 for the tri-iodide. The heat of addition per mole of ammonia
as determined from vapor pressure isotherms decreases with increase in n:
for gallium trichloride the decrease is from 32.9 kcal/mole NHa for the
monoammine to 12.6 kcal/mole NHa for the 14-ammine; for the tribromide
the decrease is from 30.3 kcal to 13.0 kcal/mole NH3; and for the tri-iodide
from 26.9 kcal to 11.4 kcal/mole NH3 (172). The compound GaC13.NH8
has m.p. 124'; dZ6 2.189 (221); b.p. 438"; AHva,, 17.7 kcal/mole and
log p,, (liq) = 7.333 - 3877/T (81).
The reaction of nitrosyl chloride with either gallium metal or gallium
trichloride yields the compound GaCla.NOC1(214). Phosphorus trichloride
and arsenic trichloride (110) form incongruently melting compounds with
gallium trichloride and Raman spectra have shown in the former case that
the fused compound is merely a solution of the dimeric gallium trichloride
in the liquid ligand (111).Consistent with this the compound has negligible
conductivity and a low heat of formation (111).By contrast, a congruently
melting compound in the system GaCl3-AsC13 can be prepared by adding
one mole of chlorine gas; colorless, hygroscopic crystals of the 1:l:l
compound are formed m.p. 5" and it is suggested, on the basis of conduc-
tivity measurements, that the structure is AsC14+GaCI4- (173). The high-
melting complex GaC13.PClsis apparently much more stable ( 1 4 0 ~ ) .
Phase studies of the systems of phosphorus oxychloride with gallium
trichloride (114) and tribromide (126) and of phosphorus oxybromide with
gallium tribromide (121) have established well-defined congruently melting
 THE CHEMISTRY OF GALLIUM 113

1: 1 compounds. Similarity in m.p. of GaCl3.POCI3and GaBr3.POC13,and

of the adducts of the trihalides with pyridine and piperidine is noteworthy.
A Raman investigation of the compound GaC13.POC13established that oxy-
gen is the donor atom (88), rather than chlorine (121).
Complexes with alkali metal halides and ammonium halides have the
general formula M+GaCl,- and have been prepared by a variety of methods
(81, 149). These include direct reaction of the univalent halide and the
gallium trihalide in a sealed tube (81), reaction in a nonaqueous solvent
such as arsenic tribromide (149) or reaction in aqueous hydrochloric acid
solution (81), though in this last procedure the product was never very
pure. That the tetrachlorogallates can be regarded as typical donor-
acceptor adducts is seen by the ease with which the chloride ion in the
complex is displaced by a strong ligand. Thus gallium trichloride is ex-
tracted by ether from lithium tetrachlorogallate and the potassium salt
reacts with gaseous ammonia a t elevated temperatures to give volatile
amines such as GaCl3.3NH3(81).
The particular case where gallium monohalide is the ligand is of interest
since this is effectively the structure of gallium dihalides Ga+GaC14- and
Ga+GaBr4-. (See Section IIJC.)
D. ELECTRICAL CONDUCTIVITY AND RELATEDPROPERTIES

The quantitative treatment of electrochemical data requires a knowl-

edge of the viscosity (q cp) and density (d gm ~ m - ~of) a fused system in
addition to its specific conductivity ( K ohm-' cm-I). The underlying theory
has been developed (103) and indicates that the degree of ionic dissociation
(a%) of a molten complex is given approximately by the expression
a N M q / d = pq
where M is the molecular weight and p the molar conductivity (MK/d).
It follows that complexes which do not dissociate into kinetically free ions
will have zero conductivity and those which are completely dissociated
into ions will have values of conductivity, viscosity, and density which give
a value of approximately 100. Some information on these properties for the
complexes of gallium trihalides is assembled in Table VIII. Choice of a
comparison temperature is difficult and various suggestions have been made
in the literature but for the data in Table VIII the melting point of each
complex has been arbitrarily chosen as the comparison temperature. (The
surface tension y dyne cm-1 is also included in Table VIII for convenience.)
It is clear from Table VIII that the trihalides themselves have a very
low electrical conductivity and this is consistent with their structure as
covalent dimeric molecules in the fused state; the degree of ionic dissocia-
tion based on these measurements is only 4-6 ppm (116, 119). By contrast
114 N. N. GREENWOOD

the dihalides are completely dissociated ionic complexes and the value of
p q approximates to 100%. The fact that values of pq somewhat above 100
are obtained can be understood in terms of a detailed analysis of the conduc-
tion mechanism for these salts, in which the cation is much smaller than
the anion (120). Complexes of the gallium trihalides fall between these
two extremes, typical values ranging from &7%. In this, the complexes
resemble those of boron trifluoride and other donor-acceptor systems (103).
The activation energies for conduction and viscous flow fall in the range
3-12 kcal/mole, which is typical of ionizing systems and for each compound
the activation energy for conduction is approximately the same as for
viscous flow, the mean value for EJE, being 1.08 (118).This is consistent
with the assumption made in developing the theory that ionic mobility
in these systems is viscosity-controlled.
Little positive information is available as to the precise nature of the
ionic species in these melts, except for the dihalides themselves, which have
been shown to ionize as Ga+GaC14- and Ga+GaBr4- (see Section 11,C).
The Raman spectrum of fused GaC13.POC13has been interpreted in terms
of the covalent adduct CLPO-tGaCL and this is undoubtedly the pre-
dominant species in the melt (88).However, the conductivity data in Table
VIII suggest that in addition there is about 4% of free ions and it is un-
likely that the Raman lines from these species will be detectable in the
presence of 99.5% of the undissociated complex, particularly as the posi-
tions of several of the lines from both structures will be similar. It is possible
that a small equilibrium concentration of ions such as POClz+and GaC14-
could be formed in this system by a ligand-switch rearrangement of the
type illustrated in the figure, though there is no direct evidence for this:
I
CI,P=O ;
kGaC1,
I/
C1,Ga ,/
// t
C1

’
A0- -=I- - PCI,
The absence of gaseous products when the molten 1 :1 and 2: 1 adducts
of gallium trichloride with pyridine and piperidine are electrolyzed elimi-
nates the possibility that the complexes ionize as H+[C6H4N-+GaC13]-,
[CaH6NH]+[CsHN-tGaC13]-, etc. Such a formulation would also imply that
the adduct with two moles of ligand would be more stable than the 1 :1
complex whereas the reverse is true. This contrasts with the adducts of
boron trifluoride where the 2 : 1 complexes are always more stable than the
1 :1 and suggests that the gallium complexes have a different structure. An
 TABLE VIII
PHYSICAL OF GALLIUMHALIDES A N D THEIRCOMPLEXES
PROPERTIES

Y 103. V d PV
Complex m.p. (dyne-cm-l) (ohm-lcm-1) (CP) (gm-cm-9 (=a%) References

GaeCIp 77.75" 27.3 1.86 x 10-3

_ _ _ _

1.823
_ ~ ~ ~

2.0536
. ~

5.8 X (115)
E
c,
Ga&? 1-22.5" 35.1 0.76 x 10-3 2.867 3.1137 4.0 x 10-4 (121) 0
I
3
Ga[GaClal 172.4" 56.2 264.2 3.355 2.4173 103.3 (120)
Ga[GaBn] 166.7" - 149.2 6.918 3.4712 136.3 (1200)
5
GaC&.POC13 118.5" 32.1 1.261 1.947 1.8304 0.442 (11.4)
%
.z
GaCkpy 126" 38.1 2.725 2.408 1,5552 1.077 (116) 4
0
GaCWpy 113" 50 9.67 2.911 1.368 6.88 (116)
GaC4.pip 134" 34.1 1.97 4.713 1.4856 1.63 (117) 0
GaCb.2pip 112" 27.3 0.620 7.571 1.3253 1.23 (117) $
GaCIrEteO 16.2" 48.5 0.810 5.77 1.8990 0.604 (106,110) E
GaBq-POBrs
GaBr3.w
154"
126"
-
42.7
1.0b
1.242
4.921b
7.090
2.8839
2.2291
1.0
1.531
(121)
(122)
9
GaBrs.pip 139" 43.8 1,503 6.844 2.1052 1.931 (122)

-The conductivities for these compounds are taken from ref. 119.
a At 161.2".

c
c
CTI
116 N. N. GREENWOOD

obvious possibility is that the 2 :1 complexes in this case are 5-coordinate

and that there is an ionization equilibrium as follows (118):
2:l Complex 1:l Complex

c1 c1 c1 c1
2L + ‘Ga’ \Ga’
/ \ / \ / \
c1 c1 c1 c1

The 2:l complex ionizes by rearrangement of the gallium from trigonal

bipyramidal to tetrahedral configuration with elimination of a chloride
ion and this ion can donate to a further GaC4 unit to give the ionic 1 :1
complex. This interprets the ionization of the 2 :1 complex and the greater
stability of the 1 : l compound. The direct formation of the 1 : l complex
from GazCleand 2 moles of ligand would involve a replacement of bridged
chlorine donors by either one or two moles of ligand to give the ionic and
covalent form of the 1 :1 complex respectively. On this basis the ionization
equilibrium of the 1 :1 complex itself (shown as the vertical equilibrium on
the right hand side of the diagram) involves a ligand replacement reaction
of precisely the same kind as that proposed for the complex GaCI3.POCl3.
E. THERMOCHEMISTRY TRIHALIDE
OF GALLIUM ADDUCTS
The energy involved in donor-acceptor reactions and the heat of forma-
tion of crystalline molecular addition compounds can be measured directly
in those cases where one of the compounds is a liquid and the other is a
solid or a liquid. The method has so far only been applied systematically
to complexes of gallium trichloride (102) but could be extended with ad-
vantage since data obtained in this way, in conjunction with heats of
sublimation, lead directly to the gas-phase heat of interaction of donors and
acceptors. The reaction sequence uses excess of the ligand as the calorimetric
liquid :
GaCL (c) + + n)L (1) = GaC18.nL(dissolved in z moles of L); -AH,,,i,
(2

+ SL (1) = GaCls-nL(dissolved in z moles of L); -AHmln

G a C h L (c)
hence,
 T H E CHEMISTRY OF GALLIUM 117

GaCl, (c) + nL (1) = GaCknL (c); -AH, = -AHmix + AH,b

It is clearly immaterial whether the solution of the complex is ionized,
dissociated, or solvated, provided only that in both reactions the same final
state is achieved.
The results are summarized in Table IX and may be illustrated by the
case of phosphorus oxychloride (104). When crystalline gallium trichloride
reacts with excess of liquid phosphorus oxychloride the heat of mixing,
-AH,i,, is 11.01 kcal per mole of GaC13. Then the crystalline 1 :1 complex
GaC13.POC13is dissolved in an excess of the same liquid, the heat of solu-
tion, -AHeoln,is 0.85 kcal/mole. By difference, the heat of formation of
the crystalline complex from crystalline acceptor and liquid ligand, - AHf,
is 10.17 kcal/mole. By measuring the heat of sublimation of the complex
(15 kcal/mole) and using the known heat of vaporization of phosphorus
oxychloride (8.4 kcal/mole), and heats of sublimation and dimerization
of gallium trichloride (8.5, 10.5 kcal/mole) the heat of the gas-phase reac-
tion, -AHr(g), was found to be:
GaCL (9) + POCb (g) = CttCla.POC1, (g); -AHr (9)= 22.6 kcal/mole

Since phosphorus oxychloride is an oxygen-atom donor in this complex

(88) this implies that the O+Ga bond in the complex is some 12 kcal
stronger than the Cl+Ga bond in the bridged dimer Ga2Cle (104).
The ligands PCL and AsCL form incongruently melting complexes with
gallium trichloride and, consistent with this, both complexes have very
low heats of formation, 3.4 and 1.4 kcal/mole respectively (109, 110, 111).
Gallium trichloride-acetyl chloride (106) also has a low heat of formation
(4.1 kcal/mole) and the difference between this value and that for the
corresponding acetone complex (15.3 kcal/mole) reflects the fact that
the negative inductive effect of the chlorine atom has reduced the elec-
tron availability at the carbonyl oxygen to a point where the oxygen
is no longer the donor atom, the structure of the complex being (4.2)
MeCO+[Cl+GaC&]-. In this case, the over-all heat of reaction not only
involves the energy involved in forming the Cl--+Gabond and the reor-
ganization energy of the gallium atom, but also the energy required to
cleave acetyl chloride into MeCO+ and C1- ions and the coulomb attraction
of the resultant ionic species MeCO+ and GaC14-.
Table I X also indicates that nitrogen donors (pyridine, piperidine)
evolve 2-3 times as much heat as oxygen donors (phosphorus oxychlo-
ride, acetone, ether) and this difference persists, though slightly less
marked, for the gas-phase reactions. Thus the heats of formation of the
gaseous complexes from gaseous ligand and monomeric gaseous GaC4
are: GaC13.POC13(g) 22.6 f 1, GaClo.EtzO(g) 22.3 f 1, GaC13.C6H6N(g)
 c1
c
m

TABLE IX
HEAT OF FORMATION OF GALLIFMHALIDECOMPLEXES
3
Complex (c or 1) -m!(kcal) Reference 1 Complex (c or 1) -AHr(kcal) Reference 3
n
GaCb-POCL (c) 10.17 k 0.01 (104) GaC&,Me&O (c) 15.3 f 0 . 9 (106) m
+
GaCbPClr (c 1) 3.40 +_ 0.03 (111) GaCL-MeCOC1(c) 4.1 fO.l (106)
M
E2
GaCLGHrN (c) 29.8 f 0 . 1 (108) GaCl..EkO 0) 9.34 f 0.08 (109)
$
GaCb-2CJLN (c) 41.5 f 0 . 1 (108) GaCL.2EkO (1) 15.48 f 0.09 (109) 0
GaCL-C5HloNH(c) 33.7 f 2 . 7 (108) GaCb.AsCla(c 1)+ 1.4 f 0 . 2 (110)
0
tr
GaClrZCJIlDNH (c) 52.1 2 . 0 (108) GaBr&HIN (c) 34.5 i - 0 . 4 (108)
 THE CHEMISTRY O F GALLIUM 119

35.2 f 2, and GaCI3C6HloNH38.1 f 2 kcal/mole. As expected, piperidine

is a stronger donor than pyridine in both the condensed and gas-phase
reactions, and those cases where both 1: 1 and 2: 1 complexes are formed
(pyridine, piperidine, ether) the second mole of ligand evolves much less
heat on addition than the first. Equilibrium vapor-pressure measurements
afford a useful check on the calorimetric data here (108) for the solid 2: 1
pyridine complex dissociates on heating to the solid 1:l complex and
pyridine vapor :
GaCla.2CsHsN (c) = G ~ C I ~ C S H (c)+
K N C S H ~ N(g).
From the dissociation pressure equation log p , , = 15.973 - 4670/T, the
heat of dissociation is -21.3 kcal/mole, and this, coupled with the heat of
vaporization of pyridine (9.7 kcal/mole) leads to a value of -AH = - 11.6
kcal/mole for the dissociation reaction
GaCb.2C6H6N(c) = GaClaCsH6N (c) + CsH6N (1).
This compares with the calorimetric value - A H = 11.7 kcal/mole for the
reverse (formation) reaction deduced from Table IX.
The final entry in Table IX refers to the complex GaBr3-CsH6N.The
heat of formation of this crystalline complex from crystalline gallium
bromide and liquid pyridine is 34.5 kcal/mole which is significantly greater
than the value of 29.8 for the corresponding chloride complex. This differ-
ence appears to persist for the gas-phase reaction (-38.3 f 2 kcal/mole)
and affords an example of the sequence I > Br > C1 > F for the heats of
formation of boron, aluminium, and gallium halide complexes (107, 108).
This order is determined mainly by the extent of ?r bonding in the mono-
meric trihalides which in turn influences the energy of reorganization of the
electron acceptor from planar to tetrahedral. (Key references to the rather
extensive literature on this point, particularly for boron halide complexes
will be found in refs. 107, 108, 184.) The reorganization energies for boron
and aluminium halides have been calculated by a simple molecular-orbital
theory (48),and, on the reasonable assumption that the trends persist with
the gallium halides, the reorganization energy of GaCL can be taken to
be 33 kcal/mole and that of GaBra 30 kcal/mole (118). This leads to the
following values for the bond dissociation energies :
D (N-GaXs) D(0-+GaCl3)
Complex (kcal/mole) Complex (kcal/mole)

GaCla.POCl8 55.6
GaCla.Etr0 55.3
-
mean 55.5
120 N. N. GREENWOOD

Since the Cl-Ga bond is some 12 kcal weaker than the 0-Ga bond
(see above) it appears that the bond dissociation energy D(C1-+GaCl3)
is approximately 43 kcal/mole.

F. CATALYTIC PROPERTIES
The early work of Ulich (279, 281) established that gallium trichloride
could be used instead of aluminium chloride in the Friedel-Crafts synthesis
of ketones and hydrocarbons. Three reactions were studied :
GaCIa
PhH + PhCOCl -+ PhrCO + HCl
PhH + PrCl--+ PrPh + HCl
PhH + C,Hd -+ EtPh + higher homologs.
The gallium catalyst, which formed a homogeneous reaction phase, was
found to promote more rapid reaction and to be less easily deactivated than
the aluminium catalyst.
In a more detailed study of the metal chloride catalyzed acetylation of
aromatic hydrocarbons with benzoyl chloride, the relative rates a t 25'
were found to be SbC16 1300, FeCL 570, GaC13 500, A1C13 1, SnCla 0.0029,
BC130.00062(152).The similarity in isomer distribution from the benzoyla-
tion of toluene argues for a common reaction intermediate, presumably
PhCOf. However, a carbonium ion mechanism is less likely for the alkyla-
tion of aromatic hydrocarbons with alkyl halides except in the case of
readily ionizable tertiary halides (29).
A detailed kinetic study of the alkylation of benzene and toluene with
ethyl bromide in the presence of gallium tribromide has shown that the
reaction is zero order with respect both to the aromatic hydrocarbon (used
as solvent) and to the ethyl bromide, and was second order with respect to
gallium tribromide (261). Toluene reacted some 24 times as rapidly as
benzene, suggesting that an aromatic complex is involved in the rate-
determining step. It is considered that, in this reaction, gallium tribromide
is a weaker catalyst than aluminium tribromide and that the second mole
of trihalide is required to facilitate the transfer of the alkyl group from the
bromide to the aromatic nucleus. The relative rates of reaction when ethyl
bromide is replaced by a series of alkyl bromides was also studied (262) as
were the o-, m-,and p-isomer distributions and partial rate factors (SO).
Solid gallium trihalides have been used to catalyze the gas-phase addi-
tion of hydrogen chloride to olefins; e.g., the reaction with ethylene a t - 20",
which is particularly well adapted to the manufacture of ethyl chloride
(213).
A continuous process for the dimerization of propylene at 200 atm
and 200-350" has been developed using, among other catalysts, trimethyl-
 THE CHEMISTRY OF GALLIUM 121

gallium or triphenylgallium (310).Similarly, ethylene is polymerized under

a variety of temperature and pressure conditions in the presence of tri-
methyl or triethyl gallium (49, 309). In the analogous polymerization of
vinyl ethers in inert solvents a t room temperature, gallium trichloride is
claimed to be a superior catalyst (126).
Gallium trichloride has also been found more effective than other metal
halides in the Lewis acid catalyzed decomposition of benzazide to phenyl-
isocyanate, and a kinetic investigation of the reaction indicates the initial
rapid formation of the 1: 1 complex which then decomposed slowly with
first-order kinetics (41):
GaCla + PhCONs + GaC13.PhCONs-+ GaClS + PliNCO + N2.
The strength of catalytic activity decreased in the sequence
GaCls > AlBrJ > AlCl3 > FeCla >> ShC16, TiCla, SnClr, TeCl,, SbBrr.
Another unusual catalytic effect which gallium trichloride has in com-
mon with many other electron-pair acceptors is the ability to enhance
greatly the reducing power of sodium borohydride towards a variety of
organic compounds (31),e.g., ethyl benzoate was rapidly reduced to benzyl
alcohol in 90% yield a t 75".
VII. Other Salts and Complexes of Gallium ( I l l )
The compounds discussed in the preceding sections have almost all
been prepared and investigated either in nonaqueous solvents or in the
absence of any solvent a t all. Aqueous systems have also been extensively
studied and the results are briefly reviewed in the present section.
Progressive dehydration of the salt Gaz(S04)3.18Hz0 leads to three new
hydrates, viz., 15, 14, and 7BB0,in contrast to aluminium sulfate which
forms hydrates containing 18, 16, 10, and 6 moles of water (278). A large
variety of other hydrated salts such as the phosphate, chlorate, bromate,
iodate, and double sulfates have been described (207, 208, 269) as well as
the hydrated perchlorate (76, 193). Earlier work on such compounds was
reviewed in 1942 (64).
A series of dithiocarbamates, (R*MCS2)3Ga,and xanthates, (ROCS&Ga,
have been prepared (53). The crystal structure of the guanidinium salt
[C(r\l'H,)3]Ga(S0,),.6H20(85) and the corresponding selenate (216) have
been reported in connection with a study of ferroelectric domains.
The analytical chemistry of gallium ferrocyanide, Ga4[Fe(CN)&has
been studied (11, 66, 272, 273) and the complex tetracyanonickelate,
Ga2[Ni(CN)4l3.6Hz0,has been prepared as a bluish-grey precipitate from
gallium sulfate and potassium tetracyanonickelate (162).
A variety of organic chelating reagents has been investigated mainly
122 N. N. GREENWOOD

from the point of view of analytical chemistry. These include substituted

quinolines (182, 183, 199, 200, 204, 219, 226), ethylenediaminetetra-acetic
acid (6429, 1,2-diaminocyclohexanetetra-aceticacid (260), acetylacetone
(36, 140, 143, l79), pyrocatechol (12, 24O), sodium 3,4-dihydroxyazoben-
zene-4'-sulfonate (MU), trihydroxyglutaric acid (67) and tartaric acid
(229, 246, 277). In addition, several tris(ma1onato) complexes have been
prepared, e.g., MIGa[CH~(CO~)2]3.3H~0 (M = Na, K) (60, S l ) , and
Tl3Ga [CHz(CO2)2]3.HzO (266).
It has been shown, further, that gallium has a much stronger tendency
to complex with ligands such as tartrate (245) and oxalate (181, 183, 244)
than has aluminium, and the species [Ga(Cz04)]+, [Ga(C20&]-, and
[Ga(Cz04)3]3- have been identified by emf, conductimetric, polarographic,
and optical density measurements. Despite this, the tris-bidentate com-
plexes are still too labile in aqueous solution to be resolvable into optically
active diastereoisomers and an earlier claim (206) to have resolved the
trisoxalatogallium(II1) anion, [Ga (C204)3lS-, has been refuted (66, 201).
Gallium acetylacetonate, m.p. 195-198" has also defied attempts to resolve
it (202).
Preliminary experiments on the potential use of liquid gallium as an
electrode in polarography have not been encouraging (91, 100, 24.9).
The use of radioactive Ga-72 in the diagnosis and treatment of bone
cancer has been investigated using organic salts such as gallium citrate and
lactate (66-68). Gallium metal itself is only mildly toxic (197).
VIII. Gallium Hydrides and Organogallium Compounds
It is doubtful whether uncoordinated gallium hydride has ever been
prepared as a stable compound and there is no evidence for its existence
as digallane Ga2HB.Earlier claims (290-292) to have prepared digallane
by the disproportionation of dimethylgallium hydride in the presence of
trimethylamine or triethylamine have been disputed (264) and a careful
series of experiments designed to liberate free gallium hydride, always
resulted in decomposition to gallium and hydrogen (264). Likewise the
preparation of an ill-defined salt-like hydride by the action of atomic
hydrogen on gallium metal (620) appears questionable (266). Apparently
a polymeric form of gallium hydride (GaH3), can be prepared by the
reaction of gallium trichloride with lithium gallium hydride in ether (297)
but the compound has not been fully characterized.
By contrast, complexes of gallane (GaH,) are well characterized and
some are moderately stable. Lithium gallium hydride was first prepared in
1947 during the pioneering researches on lithium aluminium hydride (70);
gallium chloride [or bromide (142, 295)] was reacted with an excess of
lithium hydride in ether below room temperature and, after filtration to
 THE CHEMISTRY O F GALLIUM 123

remove lithium chloride, the ether was removed by evaporation. The

compound decomposes slowly a t room temperature and more rapidly at
150" to give lithium hydride, gallium, and hydrogen (70,294). It is a milder
reducing agent that either lithium borohydride, or lithium aluminium
hydride (247, 294). Silver gallium hydride (289) and thallium gallium
hydride (296) have also been prepared a t very low temperatures but de-
compose a t -75" and -90" respectively.
The most stable adduct of gallane so far prepared is trimethylamine-
gallane, Me3N-GaH3,which is obtained by reaction of trimethylamine
hydrochloride and lithium gallium hydride, (112, 252). It melts at 70.5"
and its infrared (112) and Raman (252) spectra have been analyzed in some
detail by comparison with the spectra of trimethylamine-trideuterogallane,
m.p. 68.1" (112). Adducts with diethyl ether (295),dimethyl sulfide (854),
and several tertiary amines ( 112) have also been prepared. The etherate
reacts with hydrazoic acid below room temperature to give crystals of
gallium triazide Ga(N3)3(293).
Trimethylgallium (178), triethylgallium (54), and triphenylgallium
(92), are known and their chemistry has recently been reviewed (40).
Trimethylgallium has been prepared by the reaction of dimethyl-zinc
vapor on gallium trichloride (178) but is most conveniently synthesized
by reacting dimethylmercury with gallium metal (38, 292). It melts at
- 15.7" boils a t 55.8", has a density a t 15" of 1.151 (194), and is monomeric
in the gas phase (178). Vapor pressure measurements lead to the following
results (194):
loglo p,, (solid) = 10.5664 - 2362.2/T
log,, p,, (liquid) = 32.9398 - 2833.4/T - 8.5201 log T
AHsuhl 10.81 kcal, AH,,, (m.p.) 8.60 kcal, AHfUB
2.21 kcal,
AH,,, (b.p.) 7.39 kcal, Trouton const. 22.5 e.u.
The heat of formation of liquid trimcthylgallium was found to be 17.6 f 3
kcal/mole by bomb calorimetry (194) and 14.5 f 8 kcal/mole by reaction
calorimetry with iodine in benzene (80); the former figure is probably the
more accurate and leads to a mean bond dissociation energy, D(Ga-C), of
57.7 kcal. This is somewhat smaller than the value of 62.9 kcal for D(,41-C)
found for trimethylaluminium and this parallels the trend found for bond
dissociation energies of 4-coordinate complexes of gallium (Section VI,E)
and aluminium (118).
Less is known about the physical properties of triethyl- and triphenyl-
gallium though the former was the first organogallium compound to be
prepared (54). Triethylgallium is a somewhat viscous liquid m.p. -82.3",
d430 1.058 (54), and loglop,, = 8.224 - 2222/T; the extrapolated b.p.
124 N. N. GREENWOOD

is 142.8", AHvap10.2 kcal/mole and Trouton constant 24.5 (186).The rather

high Trouton constant may be related to the fact that, though the vapor
is monomeric (186), the liquid is somewhat viscous and the compound is
associated approximately into dimers in benzene and cyclohexane solution
(268). Triphenylgallium, m.p. 166", like the trialkyl compounds, is readily
prepared from the corresponding organomercury compound (9%').Its solu-
bility in a variety of solvents has been studied at 20" and varies from 4.15
gm/liter in heptane to 289 gm/liter in chloroform (266). The compound is
monomeric in benzene, cyclohexane, and dioxane (268).
Trimethyl- and triethyl-gallium are moderately good electron-pair
acceptors and early work established the existence of such compounds as
Me3Ga.NH3 (178), Me3Ga.NEt3 (290), Me3Ga.Etz0 (178), Et3Ga.NH3
(54), and Et3Ga.Etz0 (54). Knowledge of such adducts has been greatly
extended by the elegant work of Coates (38,39)who pointed out that, be-
cause it is monomeric, trimethylgallium is a convenient acceptor with which
to assess the relative donor properties of the elements of groups V and VI.
Trimethylgallium is a stronger acceptor than trimethylboron but is weaker
than trimethylaluminium. Despite this, weak donors which will coordinate
to trimethylgallium sometimes do not form adducts with trimethylalu-
minium because of the 20 kcal energy required to break the dimer, MeaAlz,
into monomeric Me3AI. The physical properties of the gallium adducts are
summarized in Table X. The compounds are all volatile liquids or low
melting solids and the high values of the Trouton constants can be ascribed
to the tendency for dipole association in the liquid phase and to the exten-
sive dissociation which frequently occurs in the gas phase. The gas-phase
heat of dissociation of the complexes (in kcal/mole) decreases in the
sequence (40) NMe3,21 > PMe3,18 > AsMer,10 > SbMe3 very small.
The sequence is less regular for group VI ligands, being 9.5 kcal/mole for
the dimethyl ether adduct, 10 kcal/mole for the selenium compound and
about 8 kcal/mole for the sulfur and tellurium analogs; the free energies
of dissociation follow the sequence 0 > Se > S = Te (38).
The dipole moments of triethyl- and triphenyl-gallium have been
studied in various solvents. There was zero dipole moment in heptane; the
values in benzene were 0.67 and 0.4 D, and in dioxane 2.11 and 2.14 D
(267). When compared with dipole moments of other organometallic group
I11 acceptors the relative order was B < A1 > Ga > In > T1 (267),
which agrees with conclusions drawn from dissociation-pressure measure-
ments (38).
The thermal decomposition of trimethylgallium-ammonia leads to the
formation of the dimeric compound (MeZGa.NH& and similar compounds
can be formed from the coordination compounds with methylamine and
dimethylamine (38). Reaction of trimethylgallium with methanol,
 TABLE X
COMPLEXES
OF TRIMETHYLQALLIUM

Trouton
Compound m.p. b.p. (calc) log,, pmm AHvsp(kcal/mole) constant

MeaGa.NH8 32" 179" 7.190 - 1947/T 8.9 19.7

MeaGaNHSMe 38" 169" 9.119 - 2760/T 12.6 28.6
MeaGa.NHMe2 33.6" 170" 8.596 - 2532/T 11.6 26.2
MeaGa.NMe3 96.2' 164" 7.974 - 2226/T 10.19 23.3
MeaGa-PMe, 56.7" 173" 8.851 - 2662/T 12.2 27.3
MelGaAsMe3 23.7" 121O 9.114 - 2458/T 11.2 28.5
MeaGa-SbMes 2.2" 95" 8.591 - 2105/T 9.6 26.1
MeaGaOMe2 100" 8.453 - 2078/T 9.5 25.5
MetGa.SMep 116" 9.516 - 2580/T 11.8 30.4
MeoGaSeMet -42" 111O 10.118 - 2783/T 12.7 33.0
MeaGa TeMez -32" 122" 9.172 - 2488/T 11.4 28.8
MeoGa.NCMe 25.2' 120" 8.840 - 2342/T 10.7 27.3
MeaGa.OCMez 104" 9.896 - 2648/T 12.1 32.1
Me,Ga.NEta 96" 167" 9.493 - 2909/T 13.3 30.2
126 N. N. GREENWOOD

methanethiol, methaneselenol, and acetic acid also results in elimination of

methane to give dimeric products such as (Me2Ga.SMe)*($9).These, and
similar compounds, presumably contain 4-membered heterocyclic rings of
alternate gallium and donor atoms and their properties have been studied
in some detail. For example, in some cases the ring can be split revers-
ibly by a ligand such as trimethylamine to give adducts of the type
MeSGaMez-NMerand a wide range of analogous compounds has been
prepared by this method and their properties intercompared (38,$9).
Other reactions of gallium trialkyls include their conversion by halogens
or hydrogen halides to gallium trihalides and mixed alkylgallium halides
(40, 292). Many of these form adducts with group V ligands e.g.,
MezGaC1.NHsm.p. 54" (178),MezGaC1.NMeam.p. 112.5", MezGaCl-PMeJ
m.p. 93.5", MezGaBr.NMenm.p. 136", and MezGaI.NMer m.p. 128" (40).
Dimethylgallium chloride also forms a diammoniate, m.p. 112" (178),which
has recently been shown to have the structure [Me2Ga-2NH3]+C1-(253).
Lithium aluminium hydride reacts with trimethylgallium over a period
of 24 hr at room temperature to give dimethylalane, Me2AlH,and another
product presumed to be Li+[GaH3Me]- (286), which could be considered
as a derivative of gallane. Diborane, on the other hand, reacts rapidly with
trimethylgallium at room temperature to give methyldiborane, MeBzH6,
together with gallium metal and hydrogen (2.46). This is a further illustra-
tion of the instability of uncoordinated gallium hydride discussed earlier
in this section. When the preceding reaction is carried out at -45" a com-
pound analyzing as dimethylgallium borohydride is formed as a liquid
m.p. 1.5", b.p. (extrap.) 92"; this is stable at -80" but decomposes slowly
at room temperature (246). The gas-phase molec,.dar weight corresponds
to the monomer and the compound clearly merits further examination since,
if it has the bridged structure Me2GaHzBHzit would be the first authenti-
cated example of this structural type.
Trimethyl- and triethyl-gallium are both violently hydrolyzed by water
to give compounds RzGaOH (54, 178). The reaction can be controlled by
hydrolyzing the etherate in the presence of a limited amount of water (166).
In this way, the apparently trimeric compound (Me2GaOH)3was obtained
m.p. 87", d 1.75, dipole moment 1.8 D, molecular weight 331 (trimer
requires 350) (166). The detailed X-ray analysis of this compound, however,
showed that, at least in the crystalline state, it was a cyclic tetramer
(Me2GaOH)c(260). In this form the compound has a center of inversion
and should therefore have zero dipole moment and it is possible that the
compound is trimeric in solution and tetrameric in the solid. It will be
recalled that compounds of the type (MezGaOMe)zwere cyclic dimers
(see top of this page).
In summary it can be said that organogallium compounds are highly
 THE CHEJMISTRY OF GALLIUM 127

reactive; they are violently hydrolyzed by water, and the lower trialkyls
are spontaneously inflammable. However, they are less reactive than t,heir
aluminium analogs and the acceptor power of gallium alkyls towards a
variety of ligands is less than that of aluminium. In this, the organogallium
compounds parallel the trends in acceptor properties discussed in Section
VI,E for the trihalides of gallium and aluminium (118).
REFERENCEB
1 . Adams, G. B., and Johnson, 11. L., J . A m . Ctiem. Soc. 74, 4788 (1952).
2. Akishin, P. A., Naumov, V. A., and Tatacvskii, V. M., Nauch. Doklady Vysshei
S h h l y , Khim. i Khim. Teckhnol. No. 2, 205 (1958); KristallograJiya4, 194 (1959).
3. N i l S. M., Brewer, F. M., Chadwick, J., and Garton, G., J. Znorg. &Nuclear Chem.
9, 124 (1959).
4. Apple, E. F., J. EEectrochem. Soc. 106, 251 (1958).
6. Astakhov, I<. V., Elitsur, A. G., and Nikolaev, K. M., Zhur. Obshchei Khim. 21,
1753 (1951).
6. Barrett, A. H., and Mandel, M., Phys. Reu. 109, 1572 (1958).
7. Barrer, R. M., Baynham, J. W., Bultitude, F. W., and Meier, W. M., J . Chem.
SOC.p. 195 (1959).
8. Barrow, R. F., Dodsworth, P. G., and Zecman, P. B., Proc. Phys. SOC.870, 34
(1957).
9. Bednar, J., and Smirous, K., Czeckoslov. J . Phys. 6,546 (1955).
10. Beja, hl., Nzicleo Madrid 6, No. 5, 1 (1951) also in Chirn. & ind. 67, 45 (1952).
11. Belcher, R., Nutten, A. J., and Stephen, W. I., J . Chem. SOC. p. 2438 (1952).
12. Bevillard, P., Compt. rend. 240, 1776 (1955).
IS. Bilta, W., and Voigt, A,, Z.anorg. Chein. 126,39 (1923).
14. Blakemore, J. S., Can. J . Ptiys. 96, 91 (1957).
16. Blunt, R.F., Hosler, W. R., and Frederikse, H. P. R., Pliys. Rev. 96, 576 (1954).
16. Bohm, J., and Kahan, G., 2. anorg. Chem. 238, 350 (1938).
l Y . Boisbaudran, Lecoq de, F., Compt. rend. 81, 493 (1875).
18. Bommer, H., Osterr. Chemilcer-Ztg. 44, 61 (1941).
19. Boomgard, J. van der, and Schol, K., I’hzlzps Research Repts. 12, 127 (1957).
20. Bradley, A. J., Z.Kristallog. 91,302 (1935).
21. Brandt, K., Arkiv. Kerni. Mineral. Geol. A17, No. 15, (1943).
22. Bretbque, P. de la, “Etudes Bur le Gallium,” Lausanne, Swita. Imprimeur Vau-
doise, 1955.
23. Bretbque, P. de la, Bull. Soc. Chim. France, p. 48 (1961).
24. Brewer, F. M., Chadwick, J. R., and Garton, G., J. Inorg. & Nuclear Chem. 23,
45 (1961).
26. Brewer, F. M., Garton, G., and Goodgame, D. M. L., J. Inorg. & Nuclear Chern.
9,56 (1959).
26. Brode, H., Ann. Physik 37, 344 (1940).
BY. Brown, H. C., Eddy, L. P., and Wong, R , J . Am. Chem. SOC.76, 6275 (1953).
28. Brown, H. C., Pearsall, H., and Eddy, L. P., J . Am. Chem. SOC.72, 5347 (1950).
29. Brown, H. C., Pearsall, H. W., Eddy, L. P., Wallace, W. J., Grayson, M., and
Nelson, K. L., Znd. Eng. Chem. 46, 1462 (1953).
90. Brown, H. C., and Smoot, C. R., J . Am. Chem. SOC.78, 6255 (1056).
31. Brown, H. C., and Subba Rao, B. C., J . Am. Chem. SOC. 78, 2582 (1956).
82. Brucer, M., Andrews, G. A., and Bruner, 11. D., Radiology 61, 534 (1953).
128 N. N. GREENWOOD

33. Bulewicz, E, M., and Sugden, T. M., Trans. Faraday SOC.64,g30 (1958).
34. Carlston, R. C., Griswold, E., and Kleinberg, J., J. Am. Chem. Suc. 80, 1532
(1958).
56. Casstbella, P. A,, Bray, P. J., and Barnes, R. G., J . Chem. Phys. 30, 1393 (1959).
36. Charles, R. G., and Pawlikowski, M. A., J. Phys. Chem. 62, 440 (1958).
37. Clusius, H., and Hitzig, F., Helv. Chink. Actu 33, 506 (1950).
38. Coates, G. E., J. Chem. SOC.p. 2003, (1951).
39. Coates, G. E., and Hayter, R. G., J. Chem. SOC.p. 2519 (1953).
40. Coates, G. E., “Organometallic Compounds,” 2nd ed., p. 144. Methuen, London,
1960.
41. Coleman, A., Newman, M. S., and Garrett, A. B., J . Am. Chem. SOC.76, 4534
(1954).
42. Cook, D., Can. J. Chem. 40, 480 (1962).
4s. Corbett, J. D., and Hershaft, A., J. Am. Chem. SOC.80, 1530 (1958).
44. Corbett, J. D., and McMullen, R. K., J. Am. Chem. SOC.77, 4217 (1955).
46. Corbett, J. D., and McMullen, R. K., J. Am. Chem. SOC.78, 2906 (1956).
46. Corbett, J. D., and von Winbush, S., J. Am. Chem. Soe. 77, 3964 (1955).
47. Corbett, J. D., von Winbush, S., and Albers, F. C., J. Am. Chem. SOC.79, 3020
(1957).
48. Cotton, F. A., and Leto, J. R., J. Chem. Phys. 30, 993 (1959).
49. Crawford, J. W. C., Brit. Patent, 795, 971 (1958).
60. Cunnell, F. A., Edmond, J. T., and Richards, J. L., Proc. Phys. SOC.B67, 848
(1954).
61. Davidson, A. W., and Kleinberg, J., J. Phys. Chem. 67, 571 (1953).
62. Davidson, A. W., and Jirik, F., J. Am. Chern. SOC.72, 1700 (1950).
63. Delbpine, M., Ann. chim. Paris 6, 633, 645 (1951).
64. Dennis, L. M., and Patnode, W., J. Anz. Chem. SOC.64, 182 (1932).
66. Detwiler, D. P., Phys. Rev. 97, 1575 (1955).
66. Dudley, H. C., J . Am. Chem. SOC.72, 3822 (1950).
67. Dudley, H. C., and Garzoli, R. F., J. Am. Chem. SOC.70, 3942 (1948).
68. Dudley, H. C., and Levine, M. D., J. Pharmacol. Exptl. Therap. 96, 487 (1949).
69. Durif, A., and Forraf, F., Bull. soc. franc. minbral. et crist. 81, 107 (1958).
60. Dutt, N. K., and Bose, P., J. Indian Chem. SOC.30, 431 (1953).
61. Dutt, N. K., and Bose, P., 2. anorg. Chem. 298, 131 (1958).
62. Dwyer, F. P., and Sargeson, A. M., J. Phys. Chem. 60, 1331 (1956).
63. Einecke, E., “Das Gallium,” Voss, Leipzig, 1937.
64. Einecke, E., Die Chemie 66, 40 (1942).
65. Fairhill, L. T., Ann. Rev. Med. 3, 265 (1952).
66. Fetter, N. R., and Swinehard, D. F., Anal. Chem. 28, 122 (1956).
67. Fialkov, Ya. A., and Davidenko, N. K., Zhur. Neorg. Khim. 2, 307 (1957).
68. Fialkov, Ya. A., and Fokina, Z. A., Russ. J . Inorg. Chem. 4, 1203 (1959); Ukrain.
Khim. Zhur. 26, 288 (1959).
69. Fielding, P., Fischer, G., and Mooser, E., Phys. and Chem. Solids 8, 434 (1959).
‘7’0.Finholt, A. E., Bond, A. C., and Schlesinger, H. I., J . Am. Chem. sbc. 69, 1199
(1947).
7 f . Fischer, A. G., J . Electrochem. SOC.106, 83, (1959).
72. Fischer, W., and Jubermann, O., 2. anorg. Chem. 227, 227 (1936);245,254 (1940).
73. Folberth, 0. G., Phys. and Chem. Solids 7, 295 (1958).
74. Folberth, 0. G., and Oswald, F., 2. Nuturforsch. Qa, 1050 (1954).
76. Folberth, 0. G., and Weiss, H., 2. Naturforsch. 108, 615 (1955).
 T H E CHEMISTRY OF GALLIUM 129

76. Foster, L. S., J . Am. Chem. SOC.61,3122 (1939).

77. Foster, L. S., Imrg. Syntheses 2, 26 (1946).
78. Foster, L. S., Irwrg. Syntheses 4, 111 (1953).
79. Foster, L. M., and Stumpf, H, C., J . Am. Chem. SOC.79, 1590 (1951).
80. Fowell, P. A., and Mortimer, C. T., J . Chem. SOC.p . 3734 (1958).
81. Friedman, H. L., and Taube, II.,J . Am. Chem. SOC.72, 2236 (1950).
8%.Garton, G., and Powell, H. M., J . Inorg. ck Nuclear Chem. 4, 84 (1957).
83. Gastinger, E., Angew. Chem. 67, 108 (1055).
83a. Gastinger, E., 2. Naturforsch. lob, 115 (1955).
84. Gastinger, E., Natunuissenschuften 42, 95 (1955).
85. Geller, S., and Booth, D. P., 2. Krist. 111, 117 (1959).
86. Gelsdorf, G., Muller-Heisse, H., and Schwiete, H. E., Arch. Eisenhtitlenw. 29, 513
(1958).
87. Gerding, H., Haring, H. G., and Renes, P. A., liec. trav. chim. 72, 78 (1953).
88. Gerding, H., Koningstein, J. A., and van der Worm, E. R., Spectrochim. Acta 16,
881 (1960).
89. Gevorkyan, S. V., and Garovich, N. A., Izvest. Akad. Nauk Armyan S.S.R. Ser.
Khim. Nauk 10,387 (1957).
90. Giesecke, G., and Pfister, H., Acta Cryst. 11, 369 (1958).
91. GiguEre, P. A., and Lamontagne, D., Science 120, 390 (1954).
9%. Gilmnn, H., and Jones, R. G., J . Am. Chem. SOC.62, 980 (1940).
93. Gjessing, L., Lamon, T., and Major, H., Norsk. Geol. Tidsskr. 22, 92 (1942).
94. Glazov, V. M., Mirgalovskayro, M. S., and Petrukova, L. A., IzzJest. Akad. N a u k ,
S.S.S.R. Otdel Tekh. Nauk No. 10, 68, 1957.
95. Glazov, V. M., and Petrov, D. A., Izvest. Akad. N a u k , S.S.S.R. Otdel. Tekh. Nauk
No. 4, 125, 1958.
96. Gmelin, “Handbuch der Anorganischen Chemie,” Syst. No. 36, Gallium, 1936.
97. Goldsmith, J. R., J . Geol. 6, 518 (1950).
98. Goryunova, N. A., Grigoreva, V. S., Konovalenko, €3. M., and Ryvkin, S.M.,
Zhur. Tekh. Fiz. 26, 1675 (1955).
99. Goryunova, N. A., Kotovich, V. A., and Frank-Kamenetskii, V . A., Doklady
Akad. Nauk S.S.S.R. 105, 659 (1955); through Chem. Abstr. 60, 2371 (1956).
100. Grahame, D. C., Anal. Chem. SO, 1736 (1958).
101. Greenfield, I . G., and Smith, R. L., J . Metals 7, Trans. Am. Inst. Mining Engrs.
ZOS, 351 (1955).
102. Greenwood, N . N . , J . Inorg. tk Nuclear Chem. 8, 234 (1958).
103. Greenwood, N. N., and Martin, R. L., J . Chem. SOC.p. 1427 (1953).
104. Greenwood, N. N., and Perkiris, P. G., J . Inorg. & Nuclear Chem. 4, 291 (1957).
106. Greenwood, N. N., and Perkins, P. G., Chem. SOC.Spec. Publ. No. 13, 193 (1959).
106. Greenwood, N. N., and Perkins, P. G., J . Chem. SOC.p. 350 (1960).
107. Greenwood, N. N., and Perkins, P. G., J . Chem. SOC.p. 1141 (1960).
108. Greenwood, N. N., and Perkins, P. G., J . Chem. Soc. p. 1145 (1960).
109. Greenwood, N. N., and Perkins, P. G., Pure and Applied Chern. 2, 55 (1961).
110. Greenwood, N. N., and Perkins, P. G., unpublished observations (1958).
111. Greenwood, N. N., Perkins, P. G., and Wade, K., J . Chem. SOC.p . 4345 (1957).
112. Greenwood, N. N . , Storr, A., and Wallhridge, M . G. H., Proc. Chem. SOC.p. 430
(1962).
113. Greenwood, N. N., and Wade, K., J . Chem. Soc. p. 1527 (1956).
114. Greenwood, N. N., and Wade, I<., J . Chem. SOC.p. 1516 (1957).
116. Greenwood, N . N . , and Wade, K., J . Inorg. R. Nuclear Chem. 3, 349 (1957).
130 N. N. GREENWOOD

116. Greenwood, N. N., and Wade, K., J . Chem. SOC.p. 1663 (1958).
117. Greenwood, N. N., and Wade, K., J . Chem. SOC. p. 1671 (1958).
118. Greenwood, N. N., and Wade, K., “Friedel-Crafts Reactions” (G. A. Olah, ed.)
Chapt. 6. Interscience (in press).
119. Greenwood, N. N., and Worrall, I. J., J . Inorg. & Nuclear Chem. 3, 357 (1957).
120. Greenwood, N. N., and Worrall, I. J., J . Chem. SOC.p. 1680 (1958).
121. Greenwood, N. N., and Worrall, I. J., J . Inorg. C Nuclear Chem. 6 , 34 (1958).
122. Greenwood, N. N., and Worrall, I. J., J . Chem. SOC.p. 353 (1960).
123. Greenwood, N. N., and Worrall, I. J., Inorg. Syntheses 6 , 21 (1960).
124. Greenwood, N. N., and Worrall, I. J., Znorg. Syntheses 6, 33 (1960).
126. Greenwood, N. N., and Worrall, I. J., unpublished observations (1958).
126. Grosser, F., U. S. Patent 2,457,661 (1948).
127. Gurvich, L. V., and Veits, I. V., Zzuest. Akad. Nauk, S.S.S.R. Ser. Fiz. 22, 673
(1958).
128. Hahn, €I., and Frank, G., 2.anorg. Chem. 278,333,340 (1955).
129. Hahn, H., Frank, G., Klinger, W., Storger, A. D., and Storger, G., 2. anorg. Chem.
270, 241 (1955).
130. Hahn, H., Frank, G., Meyer, A. D., and Storger, G., 2.anorg. Chem. 271, 153
(1953).
131. Hahn, H., and Juza, R., 2.anorg. Chem. 244, 111 (1940).
132. Hahn, H., and Klinger, W., 2.unorg. Chem. 260, 135 (1949).
133. Hahn, H., and de Lorent, C., 2.anorg. Chem. 270, 281 (1955).
134. Hannay, N. B. (ed.), “Semiconductom,” Am. Chem. SOC.Monograph No. 140
Reinhold, New York, 1959.
136. Hannebohn, O . , and Klemm, W., 2.unorg. Chem. 229, 337 (1936).
136. Harbeke, G., and Lauta, G., 2.Naturforsch. 118, 1015 (1957).
137. Harbeke, G., and Lautz, G., Optik 14,547 (1957).
158. Harbeke, G., and Lauta, G., 2.Naturforsch. 13a, 775 (1958).
139. Hill, V. G., Roy, R., and Osborn, E. F., J . Am. Ceram. SOC.36, 136 (1952).
140. Holm, R. H., and Cotton, F. A., J . Am. Chem. SOC.80, 5658 (1958).
140a. Holmes, R. R., J . Inorg. & Nuclear Chem. 14, 179 (1960).
141. Howell, II. G., Proc. Phys. SOC.67, 32 (1945).
14% Hutte, H. €I., British Patent 707,851 (1954); German Patent 937,823 (1956).
143. Izatt, R. M., Fernelius, W. C., Haas, G. G., and Bloch, B. P., J . Phys. Chem.
60, 170 (1955).
1&. Ivanov-Emin, B. N., and Rabovik, Ya. I., J . Gen. Chem. U.S.S.R. 14, 781 (1944).
146. Jvanov-Emin, B. N., and Rabovik, Ya. I., J . Gen. Chem. U.S.S.R. 17, 1061
(1947); through Chem. Abstr. 47, 1838 (1948).
146. Ivanov-Emin, B. N., and Rabovik, Ya. I., J . Gen. Chem. U.S.S.R. 17, 1247
(1947); through Chem. Abstr. 42, 1838 (1948).
147. Ivanov-Emin, B. N., Niselson, L. A., Rabovik, Ya. I., and Larionova, L. E.,
Russ. J . Znorg. Chem. 6, 583 (1961).
148. Ivanov-Emin, B. N., and Rabovik, Ya. I., Russ. J . Znorg. Chem. 4, 1015 (1959).
149. Jander, G., and Gunther, K., 2.anorg. Chem. 208, 241 (1959).
160. Jander, G., and Pluskal, H., 2.anorg. Chem. 201, 67 (1957).
161. Jenny, D. A., and Braunstein, R., J . Appl. Phys. 20, 596 (1958).
16% Jensen, F. R., and Brown, H. C., J . Am. Chem. SOC.80, 3039 (1958).
166. Johnson, W. C., Hammond, P., and Friedmann, H., Prog. Rept. (1947). [AS
quoted in ref. 34.1
164. Johnson, E. C., and Baskew, C. A., Inoro. Syntheses 1, 26 (1939).
 THE CHIEMISTHY OF GALLIUM 131

166. Johnson, W. C., Parsons, J. B., and Crew, M. C., J. Phys. Chem. 36, 2651 (1932).
166. Juza, R., Chemie 68, 25 (1945).
167. Juza, R., and Hahn, €I., 2. anorb. Chem. 239, 282 (1938).
168. Juza, R., and Hund, F., 2. anorg. Chem. 267, 13 (1948).
169. Juza, R., and Rabenau, A., 2. anorg. Chem. 286, 212 (1956).
160. Juza, R., and Schulz, W., 2. unorg. Chem. 276, 65 (1954).
161. Kafalas, J. A., Gabor, H. C., and Button, M. J., J . A m . Chem. SOC.79, 4260
(1957).
16.8. Karantussis, T., and Sakclarides, P., Coinpt. rend. 224, 1640 (1947).
163. Kauer, E., and Rabenau, A., Z. Naturforsch. l2a, 942 (1957).
164. Kauer, E., and Rabenau, A., Z. Xaturjorsch. Ma, 531 (1958).
166. Keith, M. L., and Roy, R., Am. Mineralogist 39, 1 (1954).
166. Kenney, M. E., and Laubengayer, ii. W., J . Qm. Chem. SOC. 76, 4839 (1954).
167. King, E. G., J. A m . Chem. SOC.80, 1799 (1958).
168. Klemm, W., and Jacobi, H., Z. arzorg. Chem. 207, 177 (1932).
169. Klemm, W., and Kilian, II., Z. anorg. Chem. 241, 93 (1939).
170. Klemm, W., and Tilk, W., 2. anorg. Cltem. 207, 161 (1932).
171. Klemm, W., and Tilk, W., 2. anorg. Chem. 207, 175 (1932).
17.8. Klemm, W., Tilk, W., and Jacobi, H., 2. anorg. Chem. 207, 187 (1932).
173. Kolditz, L., and Schmidt, W., 2. anorg. Chem. 296, 188 (1958).
174. Kolm, J. A., Katz, G., and Broder, J. D., A m . Mineralogist 42, 398 (1957).
176. Kolm, C., Kulin, S. A., and .4verbach, B. L., Phys. Rev. 108, 965 (1957).
176. Koster, W., and Thoma, B., 2. Melallk. 46, 291, 293 (1955).
177. Kotovich, V. A., and Frank-Kamenet,skii, V. A., Uchenye Zapiski, Leningrad
Gosudart. Univ. im. A. A. Zhdanoua Ser. Geol. Nauk No. 8, 135 (1957).
178. Kraus, C. A., and Toonder, F. E., I'roc. Natl. Acad. Sea. 19, 292, 298 (1933).
179. Krischen, A., and Frieser, H., Anal. Chem. 31, 923 (1959).
180. Kuznetsov, V. I., Zhur. Obshchei Khim. 20, 807 (1950).
181. Kuznetsova, V. K., Zhur. Neorg. Khim. 4, 4G (1959).
18.8. Lacroix, S., Anal. Chim. A d a 1, 260 (1947).
183. Lacroix, S., Ann. Chim. 4, 5 (1949).
184. Lappert, M. F., J . Chem. SOC.p. 542 (1962).
186. Laubengayer, A. W., and Enyle, H. R., J . A m . Chem. SOC. 61, 1210 (1939).
186. Laubengayer, A. W., and Gillian, W. I<'.,J . Am. Chem. Soc. 63, 477 (1941).
187. Laubengayer, A. W., and Schirmer, F. B., J . Am. Chem. SOC.62, 1578 (1940).
188. Laves, F., 2. Kristallog. 84, 256 (1933).
189. Laves, F., and Goldsmith, J. R., A d a Cryst. 7, 131 (1954).
190. Leifer, H. N., and Dunlop, W. C., Phys. Rev. 96, 51 (1954).
191. Levin, F. K., and Wmaus, J. W., Phys. Rev. 84, 431 (1951).
198. Linman, J. V., and Zhdanov, H. S., A d a Physicocltem. U.R.S.S. 6, 306 (1937).
193. Lloyd, D. A., and Pugh, W., J . Chem. Sor. p. 7G (1943).
194. Long, L. H., and Sackman, J. F., Trans. Fararlny Soc. 64,1797 (1958).
196. hlcMullan, R. I<.,and Corbett, J. D., J . A m . Ckem. SOC.80, 4761 (1958).
196. Margrave, J. L., J . Phys. Chem. 68, 258 (1954).
197. Meek, S. F., Harrold, G. C., and McCord, C. P., Ind. Med. 12, 7 (1943).
198. Milleron, N., 4th Natl. Symposium on Vacuum Technology, 1957, p. 38 (1958).
199. Moeller, T., and Cohen, A. J., J . Am. Chem. SOC.72, 3546 (1950).
$00. Moeller, T.,and Cohen, A. J., Anal. Chem. 22, 686 (1950).
$01. Moeller, T., and Grahn, E. H., J . Znorg. R. Nuclear Chem. 6, 53 (1957).
g0.8. Moeller, T., and Gulyaa, E., J . Znorg. & Nuclear Chem. 6, 245 (1958).
132 N. N. GREENWOOD

203. Moeller, T., and King, G. L., J. Am. Chem. SOC.74, 1355 (1952).
b03a. Moeller, T., and King, G. L., J. Phys. Colloid Chem. 64, 999 (1950).
204. Moeller, T., Pundsack, F. L., and Cohen, A. J., J . Am. Chem. SOC.76, 2615
(1954).
205. Mooney, R. C. L. Acta Cryst. 9, 728 (1956).
206. Neogi, P., and Dutt, N. K., J. Indian Chem. SOC.16, 83 (1938).
207. Neogi, P., and Mandal, K. L., J. Indian Chem. SOC.19, 67, 501 (1942).
208. Neogi, P., and Nandi, S. K., J . Indian Chem. Soc. 14, 492 (1937).
209. Neuhaus, H., Nature 180, 433 (1957).
210. Ormont, B. F., Goryunova, N. A., Ageeva, I. N., Federova, N. N., h e s t . Akad.
Nauk S.S.G.R. Ser. Fi2. 21, 133 (1957).
211. Oswald, F., Z. Naturforsch. 10a, 927 (1955).
212. Oswald, F., and Schade, R., Z. Naturjorsch. Ba, 611 (1954).
213. Padgitt, F. L., and Kirby, G. F., U. S. Patent 2,622,687 (1950).
214. Partington, J. R., and Whynes, A. L., J. Chem. SOC. p. 1952 (1948).
215. Patterson, C. S., and Tyree, S. Y., J. Am. Chem. SOC.79, 1821 (1957).
216. Pearson, G. L., and Feldmann, W. L., Phys. and Chem. Solids 9, 28 (1959).
217. Perloff, A,, J . Am. Ceram. Soc. 39, 83 (1956).
218. Pfann, W. G., U. S. Patent 2,852,351 (1958).
219. Phillips, J. P., and Duckwall, A. L., J . Am. Chem. SOC.77, 5504 (1955).
220. Pietsch, E., Seuferling, F., Roman, W., and Lehl, H., Z. Elektrochem. 39, 577
(1933).
2%1. Potter, R. F., Phys. and Chem. Solids 3, 223 (1957).
222. Powell, R. W., Nature 164, 153 (1949).
223. Powell, R. W., Proc. Roy. SOC.A209, 525 (1951).
224. Pugh, W., J . Chern. Sac. p. 1046 (1937).
226. Pugh, W., J. Chem. SOC.p. 1059 (1937).
%26. Pyatnitskii, I. V., and Kostyshina, A. P., Ukrain. Khim. Zhur. 23, 599 (1957).
227. Pyatnitskii, I. V., and Kostyshina, A. P., Zhur. Neorg. Khim. 3, 292 (1958).
288. Regel, A. R., Chem. Abstr. 62, 2488 (1958).
229. Renner, T., Z. anorg. Chem. 298, 22 (1958).
230. Richards, J. L., Nature 177, 182 (1956).
231. Ritter, E. S., J. Chem. Phys. 19, 1030 (1951).
232. Roberts, V., and Quarrington, J. E., J . Electronics 1, 152 (1955).
233. Robinson, S. R., Ann. Rept. 41, 102 (1944).
234. Roy, R., J . Am. Ceram. SOC.37, 581 (1954).
235. Roy, R., Hill, V. G., and Osborn, E. F., J. Am. Chem. SOC.74, 719 (1952).
236. Roy, R., Hill, V. G., and Oshorn, E. F., Znd. Ens. Chem. 46, 819 (1953).
837. Riidorff, W., and Landel, A., Z. anorg. Chem. 293, 327 (1957).
238. Rundle, R. E., and Corbett, J. D., J . Am. Chem. SOC.79, 757 (1957).
239. Ruff, J. K., and Tyree, S. Y., J . Am. Chem. SOC.80, 5654 (1958).
240. Ryba, O., Cifka, J., Jeskova, D., Malat, M., and Suk, V., Chem. Zisty 51, 1462
(1957).
241. Ryvkin, S. M., and Khansevarov, R. Yu., Soviet Phys.-Tech. Phys. 1, 2688 (1957).
242. Saito, K., and Terry, H., J. Chem. SOC.p. 4701 (1956).
243. Saltman, W. M., and Nachtrieb, N. H., J . Electrochem. SOC.100, 126 (1953).
644. Savchenko, G. S., Khim. Redkikh Elementov, Akad. Nauk S.S.S.R. Inst. Obshchei
i Neor6. Khim. No. 3, 57 (1957).
a&. Savchenko, G. S., and Goncharov, E. V., Zhur. Neorg. Khim. 1, 1804 (1956).
246. Schlesinger, H. I., Brown, H. C., and Schaeffer, G. W., J. Am. Chem. SOC. 65,
1786 (1943).
 T H E CHEMISTRY OF GALLIUM 133

247. Schmidt, M., and Nordwig, A,, Chem. Ber. 91,506 (1958).
248. Schubert, K., and Dorre, E.,Naturwissenschaften 40, 604 (1953).
249. Schubert, K., Dorre, E., and Giinzel, E., Naturwissenschaften 41,448 (1954).
260. Schwarzenbach, G., Gut, R., and Anderegg, G., Helv. Chim. Acta 37, 937 (1954).
261. Shafer, E. C., and Roy, R., J . A m . Ceram. Sac. 39, 330 (1956).
262. Shriver, D. F., Amster, R. L., and Taylor, R. C., J. A m . Chem. Sac. 84, 1321
(19c12).
263. Shriver, D. F., and Parry, R. W., Znorg. Chem. 1, 835 (1062).
264. Shriver, D. F., Parry, R. W., Greenwood, N. N., Storr, A., and Wallbridge,
M. G. H., Znorg. Chem. (in press).
266. Siegel, B., J . Chem. Educ. 38, 496 (1961).
266. Simmons, J. H., and Jessop, G., J . Am. Chem. Sac. 63, 1263 (1931).
267. Smirous, I<., Czeckoslov. J. Phys. 6,39 (1956).
268. Smith, R. A., “Semiconductors” Cambridge Univ. Press, London and New York,
1959.
269. Smith, F. J., and Barrow, R. I?., Trans. Faraday Sac. 64,826 (1958).
260. Smith, G. S., and Hoard, J. L., J . Am. Chern. Sac. 81, 3907 (1959).
261. Smoot, C. R., and Brown, H. C., J . A m . Chem. Sac. 78, 6245 (1956).
262. Smoot, C. R., and Brown, H. C., J . A m . Chem. Sac. 78, 6249 (1956).
263. Spandau, H., and Klsnberg, I?,, 2. anorg. Chem. 296, 300 (1958).
264. Speiser, R., and Johnson, H. L., J . r i m . Chem. Sac. 76, 1469 (1953).
$66. Srivastava, T. N., and Agrawul, S. P., J . prakt. Chem. 6, 58 (1958).
266. Strohmeier, W., and Hiimpfner, I<., Chem. Ber. 90, 2339 (1957).
267. Strohmeier, W., and Humpfner, K.,2. Elektrochem. 61, 1010 (1957).
268. Strohmeier, W., Humpfner, K., Miltenberger, K., and Seifert, F., 2. Elektrochem.
69, 537 (1959).
269. Swanson, €I. E., Gilfrich, N. ‘I‘., and Cook, M.I., Natl. Bur. Standards (U. S.)
Circular 539 p. 6 (1956).
870. Tananaev, I. V., and Bauaova, N . V., Khim. Redkikh Elementov Akad. Nauk
S.S.S.R. Inst. ObshcheZ i Neorg. Khim. No. 2, 120 (1955); through Chem. Abstr.
60,3135 (1956).
271. Tananaev, I. V., and Bausova, N. V., Khirn. Redkikh Elementov Akad. Nauk
S.S.S.R. Inst. ObshcheX i Neorg. Khim. No. 2 , 21 (1955); through Chem. Abstr.
60,3131 (1956).
272. Tananaev, I. V., and Bausova, N. V., Khim. Redkikh Elemenlov Akad. Nauk
S.S.S.R. Inst. ObshcheZ i Neorg. Khim. No.3, 41 (1957).
873. Tananaev, I. V., Glushkova, M. A,, and Seifer, G. B., Zhur. Neorg. Khim. 1, 66
(1956).
274. Tatarinova, L. I., .4uleitncr, Yu. li., and Pinsker, Z. G., KristallograJilJa 1, 537
(1956).
276. Taube, H., Friedman, H., and Wilson, A., Prog. Rept. NR 636 (1948). [As quoted
in ref. 46.1
276. Thompson, A. P., “Rare Metals Handbook” (C. A. Hampel, ed.). Reinhold,
New York, 1954.
677. Tsintsevich, E. P., Alimarin, I. P., and Marchenkova, L. F., Vestnilc Moskov.
Univ. Ser. Mat. Mekh. Astron. Fiz. Khim. 13, No. 3, 221 (1958); through Chem.
Abstr. 63, 10898 (1959).
878. Turco, A,, and Salsilli, L., Gazz. Chim. ital. 86, lW2 (1955).
$79. Ulich, H., Die Chemie 66, 37 (1943); Oel u. Kohle i n Gemeinshaft mil Brennsfofl
Chemie 39, 523 (1943).
$80. Ulich, H., and Heyne, G., 2, phys. Chem. B49, 284 (1941).
134 N. N. GREENWOOD

$81. Ulich, H., Keutmann, A., and Geierhaas, A., 2.Elektrochem. 49, 292 (1943).
888. Van Dyke, R. E., J. Am. Chem. SOC.72, 2823 (1950).
$83. Van Dyke, R. E., J . Am. Chem. SOC.73, 2829 (1950).
284. Vanleugenhaghe, C., and Pourbaix, M., CEBELCOR Rappt. Tech. NO. 74
(1958); through Chem. Abstr. 63, 5922 (1959).
886. Wagner, G. H., and Gitaen, W. H., J. Chem. Educ. 29, 162 (1952).
286‘. Wartik, T., and Schlesinger, H. J., J . Am. Chem. SOC.76, 835 (1953).
887. Welker, H., 2.Naturforsch. 7a, 744 (1952).
888. Welker, H., 2.Naturforsch. 88, 248 (1953).
889. Wiberg, E., and Henle, W., 2.Naturforsch. 7b, 576 (1952).
890. Wiberg, E., and Johannsen, T., Naturwicsenschaften 29, 320 (1941).
891. Wiberg, E., and Johannsen, T., Die Chemie 66, 38 (1942).
898. Wiberg, E., Johannsen, T., and Stecher, O., 2.amrg. Chem. 261, 114 (1943).
893. Wiberg, E., and Michaud, H., 2.Naturforsch. Qb, 502 (1954).
894. Wiberg, E., and Schmidt, M., 2.Naturforsch. 6b, 171 (1951).
896. Wiberg, E., and Schmidt, M., 2.Naturforsch. 6b, 172 (1951).
896‘. Wiberg, E., and Schmidt, M., 2.Naturforsch. Bb, 335 (1951).
897. Wiberg, E., and Schmidt, M,, 2. Naturforsch. 7b, 577 (1952).
898. Wiberg, E., and Schmidt, M., German Patent 936,387 (1955).
899. Wiberg, E., Schmidt, M., and Galinos, A. G., Angew. Chem. 66, 442 (1954).
$00. Wilson, A. S., and Taube, H., J. Am. Chem. SOC.74, 3509 (1952).
301. Wolff, G. A., Toman, L., Field, N. J., and Clark, J. C., “Semiconductors and
Phosphors,” Proc. Intern, Colloq. Garmich-Partenkirchen, p. 463 (1956).
302. Wong, R., and Brown, H. C., J. Inorg. & Nuclear Chem. 1, 402 (1955).
303. Woodward, L. A., Garton, G., and Roberts, H. L., J. Chem. SOC.p. 3723 (1956).
304. Woodward, L. A., Greenwood, N. N., Hall, J. R., and Worrall, I. J., J . Chem. SOC.
p. 1505 (1958).
306. Woodward, L. A., and Nord, A. A., J. Chem. SOC.p. 3721 (1956).
306. Woolley, J. C., and Smith, B. A., Proc. Phys. Soe. B70, 153 (1967).
307. Woolley, J. C., Smith, B. A., and Lees, D. G., PTOC. Phys. SOC.B69, 1339 (1956).
308. Yatsimirskii, K. B., J . Gen. Chem. U.S.S.R. 23, 185 (1953).
309. Ziegler, K., British Patent 713,081; U. 8. Patent 2,699,457 (1955).
310. Ziegler, K., British Patent 775,384 (1957).